Sue Hellmann - President, Product Development Ian Clark - Executive Vice President, Commercial Operations David Ebersman - Executive Vice President and Chief Financial Officer Kathee Littrell - Senior Director of Investor Relations

Joel Sendek – Lazard Capital Markets Adam Walsh - Jefferies Ian Somaiya - Thomas Weisel Partners Chris Raymond - Robert W. Baird & Co. Mark Schoenebaum - Bear Stearns Michael King - Rodman Renshaw Geoff Meacham - JP Morgan Bret Holley - CIBC World Markets Craig Parker - Lehman Brothers Michael Lever - Credit Suisse George Farmer - Wachovia Jason Kantor - RBC Capital Markets Geoffrey Porges - Sanford Bernstein May-Kin Ho - Goldman Sachs Thomas Wei - Piper Jaffray Gene Mack - HSBC Securities

At this time I would like to welcome everyone to the Genentech Q3 2006 earnings conference call. (Operator Instructions) . It is now my pleasure to introduce your host, Ms. Kathee Littrell, Senior Director of Investor Relations. Please go ahead.

Good afternoon, everyone and thank you for joining our Q3 2006 earnings call. We have posted an earnings call slide set on our website at www.gene.com that follows our earnings script. This call is being electronically recorded and is copyright by Genentech. No reproductions, retransmissions or copies of this conference call be made can be made without the written permission of Genentech. We will be making forward-looking statements, and actual results may vary materially from the statements made. Please see the Risk Factors section of our Form 10-Q for the period ending June 30, 2006 on file with the SEC for a discussion of the risk factors that could cause material variations from the forward-looking statements made during this conference call. We will be discussing financial information that includes non-GAAP financial measures in our call today. Please refer to our website at www.gene.com under the Investor tab, and click on Financials for the most directly comparable GAAP financial measures, with a reconciliation to the non-GAAP financial measures that we will be discussing today. Today I'm joined by Ian Clark, our Executive Vice President of Commercial Operations; Dr. Sue Hellmann, President of Product Development; and David Ebersman, our Executive Vice President and Chief Financial Officer. Now I will turn the call over to Ian for opening comments.

Good afternoon everybody. Total U.S. sales in Q3 were $1.83 billion, an increase of 34% from the $1.7 billion during the same quarter last year. Today, as always, I will be providing information on approved and unapproved uses. I want to remind you that Genentech policies only allow our sales force to promote products for an on-label use. Let me start with a brief update on the changes to our distribution model. As a reminder, beginning in Q3, we changed our product distribution model for Avastin, Herceptin, Rituxan, and these changes have been fully implemented. Overall, customers accepted these changes well. As I said, we do not believe the distribution changes had a material impact on Q3 sales. The use of third party vendor days reporting from both IMS and NVC have been somewhat unreliable. Distribution changes, as well as changes in business practice relationships, often have unexpected implications on their data collection process. We continue to work with both IMS and NVC as appropriate to identify potential sources of data inaccuracies. We want to remind you that this data does not come from us and you should use this data at your own risk. Moving on to our products, beginning with oncology. First Avastin. Avastin U.S. sales in Q3 were $435 million, an increase of 34% from the $325 million during the same quarter last year. Year-over-year growth resulted primarily from increased use in metastatic non-small cell lung and breast cancers, both unapproved uses of Avastin. There were also modest gains in metastatic colorectal cancer, where Avastin is approved in the first and the second line settings. These were partially offset by declining revenues in metastatic renal cell cancer and late stage pancreatic cancer, both unapproved uses. In metastatic colorectal cancer, opportunities to further increase penetration duration in first line are limited. In the second line we believe opportunity remains to drive increased penetration in addition to the use of the label pending, I think, every two weeks dose. In non-small cell lung cancer compared to last year we have observed an increase in adoption at a large number of physicians using Avastin. Considerable opportunity remains for additional growth in the market once we receive the final leg from the FDA based on the 84599 study. We have finalized our launch preparations, including training our sales force, and on repromoting Avastin in this setting immediately upon approval. In metastatic breast cancer adoption has increased over last year, however, we have observed a recent slowing. Despite compendia listing of Avastin in NVC, physicians’ concern about reimbursement limits adoption in this setting. Moving on to Herceptin. U.S. sales of Herceptin in Q3 were $302 million, a 40% increase compared to the $215 million during the same quarter last year. Herceptin sales in HER2-positive metastatic breast cancer continue to be strong, driven by slight increases in duration relative to last year. Given the adjuvant setting, an unapproved use has also increased relative to last year; however, compared to last quarter, use remained steady and has possibly reached a plateau. As you are aware, our original action date of August 17 of this year was delayed due to the FDA's request for additional analyses of previously submitted studies. The FDA review period was extended by 90 days, and as a result we are preparing for a potential Q4 launch for this indication. Next Tarceva. Third quarter U.S. sales of Tarceva were $100 million, up 37% from the $73 million received in the same quarter last year. Genentech market research in Q3 indicate an increase in front line pancreatic cancer penetration, but it decreased in second line non-small cell lung cancer use as compared to Q2 this year. Now on to Rituxan. Q3 U.S. sales for Rituxan reached $509 million, up 12% from the $456 million in the same quarter last year. Rituxan's overall adoption rate and combined market in NHL and CLL, including unapproved uses remain steady at 82% in Q3 of this year. On December 29 this year the FDA approved Rituxan for two new indications. First, for the use of Rituxan in the first line treatment for follicular NHL in combination with CVP therapy. The second is the use in the treatment of low-grade NHL patients with stable disease, all those who achieve partial complete response following first line CVP therapy. The sales force is now promoting Rituxan in both these settings. These two new indications, along with the approval received in February for diffused large B-cell lymphoma, further increases Rituxan on-label use in NHL. This brings our total approved uses in NHL to four. As you know, the use of Rituxan in three out of four of these patient populations is already fairly high. Moving on to the immunology space, we continue with Rituxan in RA. The launch is going well as it meets our expectations. Our promotional efforts continue to be competitive relative to the other biologics in the RA marketplace. Market research indicates that 60% of our physician targets have used Rituxan for RA, and over half use Rituxan in multiple patients. In total approximately 10,000 patients had been reached with Rituxan in the immunology setting so far this year, with 70% of use coming after patients have inadequate response to two or more anti-TNF therapies. As we stated previously, there are significant hurdles to reliably measuring the portion of that Rituxan sales to do with RA. Next slide, immunology products. U.S. sales of Xolair were $107 million in Q3, a 30% increase compared to the $82 million into the same quarter last year. And U.S. sales for Raptiva were $23 million in Q3, up 10% from the $21 million in the second quarter of last year. Now on to our tissue, blood and repair products, and beginning with LUCENTIS. U.S. sales in LUCENTIS in Q3, its first full quarter in the market, were $153 million compared with $10 million in Q2 last year. Initial indicators project the launch is going very well. Nine weeks after the launch market research showed that LUCENTIS achieved approximately 15% share of newly diagnosed patients. We continue to monitor dosing practices and observe that they vary widely. Early market research suggests that many rectal specialists will follow the schedule recommended in the label, dosing each patient monthly for the first four months before they consider adjusting the long-term dose and frequency for individual patients. Moving on to reimbursement. Despite broad coverage, timely payment to early adopters remains an issue, and of course some offices carry significant accounts receivable. All Medicaid carriers have indicated they will cover LUCENTIS, and most have now begun to pay submitted claims while the payers have begun to follow their directions. The reimbursement issue should continue to improve over time as more physicians receive payment for their initial LUCENTIS claims and gain competence in its smooth reimbursement process. Market research also suggests that while we have seen rapid adoption of LUCENTIS, the total market has not grown. Increasing the LUCENTIS market share to date has come at the expense of other agents, most notably Macugen, but also from Visudyne and Avastin. Approximately 80% of LUCENTIS patients were existing patients who have switched to LUCENTIS, and the remaining were newly diagnosed. Over the course of the next several years the existing patient pool will diminish, and the sales for LUCENTIS will rely more heavily upon acquiring and retaining newly diagnosed patients. Initial data also suggests that off-label use of Avastin will remain a factor in the marketplace. Now on to our other tissue growth and repair products. U.S. sales in Nutropin were $92 million in Q3, up 3% from the $89 million in the same quarter last year. U.S. sales of Thrombolytics products were $60 million in Q3, up 3% from $58 million in the last year. And finally U.S. sales of Pulmozyme were $50 million in Q3, That is a 6% increase compared to $47 million in the same quarter last year. Now I will turn the call over to Sue.

Thank you, Ian and good afternoon, everyone. This quarter has been significant for our BioOncology and immunology product development programs. Starting in BioOncology, as Ian mentioned, we received two additional indications for Rituxan in non-Hodgkin's lymphoma. We're also on track for tomorrow's PUDFA date for Avastin in combination with carboplatin and paclitaxel for first line non-squamous, non-small cell lung cancer. And we're working with the FDA to finalize the wording of the potential label for the sBLA for Herceptin as early stage or adjuvant treatment of HER2-positive breast cancer. We expect to hear from the FDA by November 16. In immunology, the top line results of the Phase II Rituxan and in relapsing remitting multiple sclerosis study exceeded our expectations. These results have been extremely encouraging as we move forward with our plans to expand our immunology franchise and further investigate the role of B-cell therapies as treatment for the vast number of patients who struggle with autoimmune diseases. We also moved forward in our efforts to grow our early pipeline by adding two new small molecule entities, the PARP inhibitor for malignant melanoma through our collaboration with Inotek, and a systemic hedgehog antagonist for cancer through our collaboration with Curis, as well as two unnamed new molecular entities for oncology. These new additions bring us to a total of five new molecular entities added to our pipeline so far this year, exceeding our goal of adding four entities for our development pipeline in 2006. We and our collaborators also initiated enrollment in the Phase III Tarceva in adjuvant non-small cell lung cancer study called RADIANT and the Phase III TNKase Catheter Clearance study. We’ve completed enrollment in the Phase III Avastin adjuvant colon cancer study, NSABP-C08, the Phase III Roche study evaluating two doses of Avastin in non-squamous, non-small cell lung cancer called AVAIL, and the Phase II gemcitabine plus/ minus Omnitarg study in advanced ovarian cancer. Now let me turn to an overview of Avastin. As you know, we received a FDA Complete Response Letter for the sBLA submission of Avastin in combination with chemotherapy for first-line metastatic breast cancer. We will work closer with the FDA as we prepare for a resubmission of the sBLA currently anticipated by mid-2007. We will be conducting meetings with the FDA for future cooperative group studies to ensure agreement regarding the study protocols. We are in the process of prioritizing and scheduling these meetings for upcoming studies. We believe in the long run these meetings will optimize our ability to partner with groups and meet the evolving standards of the FDA. We will communicate updates to our study timelines that result from these meetings. Enrollment in the U.S. has been slower than expected in the metastatic breast cancer study known as Ribbon 1 and Ribbon 2. However, we have now launched global enrollment in Ribbon 1, which we believe will help encourage enrollment. We're also exploring the possibility of global enrollment in Ribbon 2. Roche's first-line metastatic breast cancer study, AVADO evaluating the 7.5 versus 15 mg per kg doses of Avastin is continuing to enroll patients. Later this year we will review the primary safety information from ECOG 2104, our pilot adjuvant breast cancer study, which may enable us to finalize our plans for the Phase III program in adjuvant breast cancer. The timeline and study design for the adjuvant breast cancer study will be provided following the FDA meeting to assess that safety information from ECOG 2104. As we prepare for the launch of Avastin in first-line non-small cell lung cancer, we continue our investigation of the efficacy of Avastin in earlier stage disease. We have our FDA meeting scheduled for Q4 '06 for the adjuvant non-small cell lung cancer study, ECOG 1505. Additionally, Roche anticipates results from AVAIL, their non-small cell lung cancer study investigating the 7.5 versus 15 mg per kg dose of Avastin in combination with gemcitabine/cisplatin chemotherapy in the first half of 2007. In the first-line colorectal cancer study results from Roche's XELOX or FOLFOX plus or minus Avastin study 16966 were presented at ESMO this month. The study met its two primary endpoints in that non-inferiority of XELOX to FOLFOX was established, and further evidence of the progression-free survival benefits of adding Avastin for oxaliplatin-based chemotherapy was provided. However, the magnitude of benefit was lower than we had seen in previous studies and the duration of Avastin therapy was shorter than that of our previous clinical trials. Our interpretation of this data suggests the importance of treatment to progression. Further analyses are being conducted and will be presented at the GI-ASCO in January of 2007. The Avastin Phase III adjuvant colon cancer study, NSABP-c-08 completed enrollment in October. Regular interim analysis will take place for this study beginning sometime in 2007. Timing of final results will depend on the magnitude of the difference between the two treatment arms. Enrollment in Roche's adjuvant lung cancer study called AVANT is going well since it was reopened. This study is expected to be fully enrolled by the first half of 2007. CALGB 90206, our Avastin renal cell cancer study, is an event-driven trial and is fully enrolled. We've not received a recent update from CALGB. We continue to look forward to the results from this study and from Roche's Avastin renal cancer study, [AVORIN], as the data from these trials matures. Our first-line ovarian cancer study, GOG218, is being conducted under a special protocol assessment and is enrolling patients slowly. The second line study, GOG213, is awaiting a meeting with the FDA to discuss the protocol. Regarding Herceptin, we plan to submit two separate sBLAs to the potential adjuvant HER2-positive breast cancer label by early 2007. The first of these will be based on one-year data from HERA, and the second one on BCIRG 006. These studies evaluated Herceptin in HER2-positive patients who are high-risk lymph node negative, as well as Herceptin administration every three weeks. Since submissions could potentially also expand treatment options in the HERA regime provided Herceptin following adjuvant chemotherapy and the BCIRG study included in our investigating Taxotere, Carboplatin and Herceptin. Turning to Tarceva, OSI initiated enrollment in their adjuvant non-small cell lung cancer study in EGFR-position patients called RADIANT in Q3 of this year. The objective of this study will be to examine whether the addition of daily Tarceva following standard treatment of early-stage non-small cell lung cancer will improve the D3 survival in approximately 900 patients. The first line non-small cell lung cancer study, SATURN, continues to enroll patients, as does TITAN, the second line head-to-head comparison of Tarceva to chemotherapy. Regarding Rituxan in the oncology setting, we expect to complete enrollment in our Phase III label enabling Rituxan study in relapsed chronic lymphocytic leukemia, known as REACH in 2007. This global study is being led by Roche and compares fludarabine cyclophosphamide and chemotherapy to this chemotherapy combined with Rituxan. The primary endpoint is progression-free survival. A clinically important endpoint is CLL]. We have decided to collaborate with Roche on the PRIMA study, a 1,200 patient study evaluating physicians' choice of three chemotherapies plus Rituxan followed by a maintenance regimen of Rituxan every eight weeks over 24 months versus observation in patients with first-line follicular non-Hodgkin's lymphoma. We anticipate enrollment will be complete by Q1 2007. Now turning to Rituxan in the immunology setting, there will be four oral presentations reviewing Rituxan data at the annual meeting of the American College of Rheumatology in November. One focus is on inhibition of joint erosion in Rituxan-treated patients and discussing subsequent treatment or combination therapy with Rituxan. Our strong presence at ACR is evidence of the growing recognition of Rituxan's potential to play an important role in the treatment of RA, as well as the amount of ongoing research investigating the safety and efficacy of additional courses as to the long-term safety of Rituxan. Our extensive DMARD inadequate responder or IR program, continues to enroll patients. We anticipate that enrollment will be completed in the global Phase III methotrexate IR study, SERENE, and the TNF-IR controlled retreatment study, SUNRISE, by the end of this year. The methotrexate IR study has a six-month endpoint, while the controlled retreatment study follows patients for 48 weeks. The top line results of Phase II RRMS Rituxan study exceeded our expectations in meeting the primary endpoint, a reduction in the total number of Gadolinium-enhancing T1 lesions, and also shows a statistically significant reduction in the number of relapses among Rituxan-treated patients. These results are impressive in a six-month timeframe, although we believe it is important to examine whether they will hold up in a larger study and over the course of time. The teams are still considering next steps, and we plan to make a decision in the fourth quarter of this year. We're still awaiting results of the 96-week endpoint in the PPMS study, which completed enrollment in the final quarter of 2005. Now turning to the second generation humanized anti-CD20 molecule, results from the Phase II study in RA will be presented at the American College of Rheumatology in November. We continue our work to develop plans for three Phase III humanized anti-CD20 RA studies, one investigating methotrexate IR patients, one TNF/IR patient, and a third x-ray study. We expect to initiate these trials in the first half of 2007. Given the positive top line result of the RRMS Phase II study we are evaluating our immunology, neurology program, including NMO. Now I would like to provide some brief comments about our disagreement with Biogen Idec. We continue to pursue a resolution of our differences regarding our 2003 collaboration agreement covering the humanized anti-CD20 product. The disagreement is around the details of decision-making rights. We believe Genentech can proceed with further trials with humanized anti-CD20 molecules based on existing agreements between the companies, and Biogen Idec degrees with our position. These disputes escalate as contemplated by the 2003 contract and are not a topic of arbitration. There is a risk that the resolution of an arbitration would require that both parties agree to development decisions before moving forward with the humanized anti-CD20 molecule trials, in which case we might have to alter or cancel planned trials in order to get Biogen Idec's approval. Now turning to LUCENTIS. We are pleased with the retinal specialist and patients to LUCENTIS. We have now shifted our focus to exploring potential new ophthalmic indications and opportunities. We are in discussions with the FDA regarding the design of a pivotal trial evaluating LUCENTIS as treatment for diabetic macular edema, and we're planning for a retinal vein occlusion study. We will communicate more about the timelines of these potential studies as the plans progress. The safety and utility of pre-filled syringes are also being examined in our LUCENTIS extensive study, HORIZON, starting this month. We have extended this study for an additional three years, which will allow us to gather up to five years of data on treatment and safety for LUCENTIS-treated patients. An interim analysis of 500 patients from HORIZON will be presented at the American Academy of Ophthalmology in November. We continue our efforts to further understand the optimal treatment regimen for LUCENTIS. In the second half of 2007 we anticipate results from the first cohort of SAILOR, the Phase IIIb study in which approximately 2,400 patients will be followed for 12 months using criteria-based retreatment. We also anticipate the two-year ANCHOR study results in the first of 2007. The LUCENTIS Phase II RESOLVE study being conducted by Novartis continues to enroll diabetic macular edema patients. We're aware that National Eye Institute and the NIH have indicated plans to fund a new multicenter clinical trial to compare LUCENTIS and Avastin in the treatment of wet AMD. Genentech is not taking part in the planning or the implementation of this potential study, as we feel that LUCENTIS is the right drug for the treatment of wet AMD, and that our rigorous clinical trial program has supported that view. Our mission is to meet significant unmet medical needs through innovation and we believe the greatest benefit we can provide to patients is to focus our continued development of LUCENTIS to treat diseases such as diabetic macular edema and retinal vein occlusion. Finally, turning to our early development programs, we're collaborating with Inotek on their ongoing Phase Ib trial investigating the PARP inhibitor as a potential treatment for malignant melanoma in combination with temozolomide. PARP is an enzyme within cells that directs the repair of damaged DNA and has applications for the treatment of cancer and cardiovascular conditions. In those models of cancer PARP inhibition enhances the ability of therapy to kill tumor cells. In Q3 we submitted an IND for Systemic Hedgehog Antagonist Therapy for cancer treatment, and we now expect to submit an IND anti-BR3 for CLL in 2007. Let me finish up by highlighting a few upcoming milestones for Q4 2006. We expect to complete enrollment in two Phase III Rituxan RA studies. SERENE, the methotrexate IR study and SUNRISE, the TNF-IR controlled retreatment study, and the Apo2Ligand/TRAIL Phase I(a) study for solid tumors and hemotalogic malignancies. We also expect to make a Phase II go/no go decision for BR3-FC in RA and a Phase II go/no go decision in our B-cell program for relapsing remitting MS. Now I will turn the call over to David.

Thank you, Sue and good afternoon. I will start off with an update on manufacturing. We are pleased to report that our capacity enhancement projects continue to be on track, and in Q3 2006 we achieved another important milestone when the FDA approved Wyeth to manufacture bulk Herceptin at their Andover, Massachusetts facility. Wyeth will produce approximately 25% of our Herceptin requirements for the next several years, and it joins Lonza and Novartis as key third-party sites we have licensed to produce our products. We remain on track for our remaining 2006 manufacturing milestones, including licensure of another 20,000 liters of capacity to manufacture bulk Avastin at our Porrino site, and the FDA approvals for yield improvement projects on Rituxan and Avastin. We continue to anticipate the licensure of 90,000 liters of capacity at our Oceanside facility, NIMO, for the production of both Avastin in the first half of 2007. Although manufacturing remains at risk for our business, we are pleased with the progress we've made to increase capacity, and we continue to anticipate that we may be able to reduce our overall capacity utilization down somewhat from 100% in 2007, providing us with a modest buffer against unexpected future production problems. Moving on to a legal update, as you probably read, last week the U.S. Supreme Court heard oral arguments in our lawsuit with MedImmune regarding the Cabilly patent. The issue at hand in the Supreme Court is not the validity of the patent, but is whether or not MedImmune has the right to sue us in an attempt to challenge the patent, while simultaneously enjoying the benefits of a license for the patent. If the Supreme Court reverses the lower courts who ruled in our favor on this matter, the court may find that MedImmune has standing to challenge the Cabilly patent in court. We expect a Supreme Court ruling no later than the second quarter of 2007. Turning now to the financials, as Kathee mentioned earlier, unless otherwise noted my comments on expenses and income items are on a non-GAAP basis, which exclude the impact of recurring charges related to the 1999 Roche redemption, litigation-related special items, and employee stock-based compensation expense. I will start now with the revenue component of the income statement, adding to the U.S. product sales information already noted by Ian. Sales to collaborators were $111 million this quarter, a 29% increase over Q3 of last year, primarily due to higher sales of Herceptin to Roche. For the full year 2006 we continue to expect sales to collaborators to increase approximately 40% year-over-year. Note that sales to collaborators can fluctuate on a quarter by quarter basis depending on collaborator ordering patterns. Royalty revenues were $364 million in Q3, a 53% increase over Q3 of last year, substantially driven by higher royalties from Roche, which represented 63% of Q3 2006 overall royalties. For the full year 2006 we're now expecting increasing sales by Roche to drive royalties up by about 45% over 2005. Contract revenues were $79 million this quarter, a 25% increase over Q3 of last year. For the full year 2006 we expect to see a slight increase in contract revenues over 2005 levels. Our expectation for Q4 is that contract revenues will be significantly lower than the first three quarters of 2006 due to normal fluctuations in this line, such as the fact that we do not expect to receive any significant milestone payments during the fourth quarter. Total operating revenues were approximately $2.4 billion in Q3 of this year, a 36% increase over Q3 2005. Turning now to the expense line items. Cost of sales was $297 million, or 15% of net product sales in Q3 this year, a decrease from 16% of net sales in Q3 2005, primarily due to a favorable product sales mix, including LUCENTIS and higher BioOncology sales. We continue to expect cost of sales as a percentage of product sales to be approximately 16% for the full year 2006. We will caution that cost of sales can always come in higher than expected if we have unforeseen manufacturing or inventory issues. Non-GAAP R&D expenses were $419 million this quarter, a 27% increase over Q3 2005, driven by higher licensing deals this quarter, increased development spending on late stage clinical trials, and higher research headcount and activity. R&D was 18% of operating revenues this quarter, a decrease from 19% in Q3 2005, driven largely by higher operating revenues. Consistent with prior years, we expect R&D spending in the fourth quarter to trend higher, with much of the forecasted increase targeted for incremental in-licensing investments. This year in particular we continue to aggressively seek out new deals for products and technologies that can help drive our future growth, such as the deals we completed with Inotech and CGI in the third quarter. As you know, there is always inherit unpredictability of completing new deals, so it is difficult to pin down our expected in-licensing spend in the fourth quarter. In Q4 we're also planning an increase in clinical manufacturing campaigns that will impact R&D expense. Our current thinking for the full year 2006 is that we expect R&D to be towards the lower end of 18% to 19% range of operating revenues. Non-GAAP MG&A expenses were $460 million this quarter, a 34% increase over Q3 2005, largely due to increased marketing and sales spending primarily in support of our recent product launches and incremental co-pay assistance donations this quarter. MG&A as a percentage of operating revenues was 19% this quarter, a decrease from 20% in Q3 2005, driven largely by higher operating revenues. We're now expecting MG&A as a percentage of operating revenues for the full year 2006 to be approximately 20%. Collaborative profit-sharing expenses were $250 million this quarter, an increase of 14% over Q3 of last year, primarily due to higher sales of Rituxan, Xolair and Tarceva. Non-GAAP pre-tax operating margin as a percentage of total revenues was 40% this quarter, an increase from 36% in Q3 2005. For the full year 2006 we're now expecting our operating margin to be about 38% as compared to 32% in 2005. Other income net was $55 million this quarter, more than doubling from the Q3 2005 number, driven by higher investment income and gains from certain of our biotech equity investments. For the full year 2006 other income net is expected to be about 2.5 times the 2005 level of $91 million. This increase is driven by the biotech stock gains we recognized last quarter and higher interest income. On taxes, our non-GAAP tax rate was 37% in Q3 of this year, a decrease from 41% in Q3 2005. We currently expect our tax rate for the full year of 2006 to be about 38%. Non-GAAP net income this quarter was $637 million, or $0.59 per share, a 66% increase in net income and 69% increase the EPS over Q3 last year. Weighted average shares outstanding for computing diluted non-GAAP EPS were 1.072 billion shares this quarter, compared to 1.087 billion shares in Q3 of last year. Now turning to some cash metrics, our unrestricted cash and investments portfolio totaled approximately $4 billion at September 30, 2006 as compared to $4.2 billion at September 30, 2005. Cash from operations in Q3 2006 was $421 million, and cash used for capital expenditures was $350 million. CapEx for the full year 2006 is now expected to be approximately $1.3 billion. In the third quarter of 2006 we spent $218 million for gross share repurchases, with a net effect on our cash position of negative $67 million, factoring in the offsetting cash inflows to the Company from stock option exercises and the tax benefits related to those option exercises. We plan to continue to repurchase shares in 2006 to offset dilution from option exercises and to maintain Roche's ownership percentage above 55.7%. Our days of sales outstanding with respect to product sales accounts receivable was 37 days at the end of this quarter compared to 32 days in Q3 2005. The increase in the average collection period is largely attributable to the extended payment terms offered to certain wholesalers in conjunction with our launch of LUCENTIS on June 30. Turning now to stock option expensing, employee stock-based compensation expense for Q3 2006 was approximately $76 million on a pre-tax basis, or $46 million after taxes, or approximately $0.04 per share. We allocated $35 million to the R&D line, $41 million to the MG&A line, and recognized a tax benefit of $30 million in the income tax line related to employee stock-based compensation expense. Now that we have nearly a full year of actual stock activity under both the stock option and employee stock purchase plans, we currently expect that stock-based compensation expense for the full year 2006 will be in the range of $0.17 to $0.18 per share. In terms of new stock option grants awarded to employees of 2006, we expect the total grants to equal approximately 1.4% of our shares outstanding, which is modestly lower than what we communicated previously. The bottom line for 2006, recognizing that there are a number of factors that can affect our forecast, we have revised our current expectation for year-over-year non-GAAP EPS growth to a range of 65% to 70%. This change from our previously communicated expectations is based primarily on higher 2006 forecast for LUCENTIS sales, higher forecast for royalties, and our current thinking regarding incremental investments in the business. Let me wrap up my comments by saying it was an eventful third quarter, highlighted on the positive side by the very successful launch of LUCENTIS, and on the negative side by the Complete Response Letter from the Avastin breast cancer trial. Overall, we believe we're on track to meet almost all of the key 2006 goals we set for the business at the beginning of the year and as always, we remain enthusiastic about executing on the business opportunities immediately in front of us while maintaining our commitment to scientific excellence and our long-term focus on translating our scientific discoveries into innovative products that provide meaningful benefits to patients and create value for our shareholders. Now, I will turn the call back over to Kathee.

As per our usual practice, I would like to ask that each of you ask one question and one question only, so that as many of you as possible can get your questions answered. Operator, can you please queue up the Q&A session?

(Operator Instructions) Your first question comes from Joel Sendek – Lazard Capital Markets. Joel Sendek – Lazard Capital Markets: Hi, thanks. I have a question about the distribution change. If you can give us any more detail or quantify the sales impact that it had. You mentioned before about one-day sales, is that how much it was? Did this have a different impact on the three drugs that it affected? Thanks.

There were a couple of things which we have mentioned in the past. The first one was whether or not there was any anticipatory stock changing at the end user stage, hospitals, and the second was the one-day's worth of sales. There may have been a bit of the former, but as I said in my prepared comments, nothing which we thought was material. The one day was actually a Friday, and we sell very little on a Friday, so that was relatively small. So, we feel it was a correct comment that the impact was not material at this point.

Our next question comes from Adam Walsh - Jefferies. Adam Walsh - Jefferies: Hi, thanks. Ian, you mentioned that off-label uses for Avastin in breast and lung cancer have been somewhat tempered by compendia listing and physician concerns regarding reimbursement under that system. To what extent should we expect formal FDA approval of Avastin, say in lung cancer, to impact overall adoption? How should we think about the magnitude there? Thanks.

Just to clarify the earlier comment, I think we have seen somewhat different dynamics. The use of Avastin in lung has continued to be consistent; no real change in that and I will come to your next question in a second. It is the breast cancer where we’ve seen some jerkiness. It took a while to get the compendia listing. We know it is going to be relatively expensive, and of course there was a delay in the label, so we have see some changes there. As I have said in previous calls we are experiencing considerably less off-label adoption these days than we did several years ago, even after compendia listing. There will always be that impact of what we sell in the short-term, it generally means that the upside that we will see when we get a label is fairly considerable. We have not quantified that, but I think we do expect a considerable upside. Hopefully we get the label tomorrow.

Our next question comes from Ian Somaiya - Thomas Weisel Partners. Ian Somaiya - Thomas Weisel Partners: Thanks for taking my question. Just related to LUCENTIS, I wanted to follow-up on your comments that sales obviously came at the hands of other drugs and the market is not growing any faster. How should we think about LUCENTIS going forward? Is there incremental growth above and beyond what we have seen in 3Q, and how much growth is there?

Let me try and give you three or four facts to think through. Clearly, the long-term penetration in the marketplace is going to be pretty essential. I don't doubt that that will be fairly high. I tried to give you some thoughts around how the dynamic will play out between existing patients switching to LUCENTIS and new patients. I think what we are seeing at launch is a lot of switch to the proportion of new. Eventually we will run out of switch, and you have to think about how that plays out. The factor of dose is going to be fairly critical. We are in very much the early stage of product lifecycle. The label suggests you should dose every month. I would suggest that is what has happened. They have options beyond the first four months. The doses may get less frequent. Then there is the longer-term issue about how it is used chronically or not. In terms of market growth, I think one could speculate, given the quality of the drug, that you might see more patients being treated, but it is a pretty early stage to get into that.

Our next question comes from Chris Raymond - Robert W. Baird & Co. Chris Raymond - Robert W. Baird & Co.: Thanks for taking the question. Ian, you mentioned that the distribution channel switches. It happened again in the third quarter and was not the driver of Rituxan and specifically Herceptin being down sequentially. I was wondering if you could maybe talk to maybe a couple of reasons for why they would be, though?

Let me start off by talking about Rituxan a bit. Rituxan is a big drug. It is $2 billion, $2 billion-ish. It has not been growing at 30% for many years, and that is largely due to the fact that our penetration in many of the hemo segments is pretty high. I don't doubt that the new labels will help with that, but we didn't have those labels during the quarter. The RA launch is on track, but you are going to have to accept that TNF IR marketplace is relatively small. It is not going to add $500 million, $1 billion overnight. We may see some additive impact later on this year with the RA going and the new label, but that didn't necessarily fly in the last quarter. The final thing I would draw your attention to is that quarter-on-quarter growth for Rituxan has been a little staccato over the years. Nearly always when we have had a strong quarter it has either been followed by or preceded by a relatively weak quarter. Actually the 10% growth that we saw between Q1 and Q2 was actually the highest quarter growth that we'd seen for about four years. One might have expected the following quarter may not have been quite so great. So there's a number of factors you might want to think about there. Chris Raymond - Robert W. Baird & Co.: What about Herceptin?

We always knew that the off-label adoption -- because that is where we are at in the adjuvant setting -- was likely to reach some sort of plateau. It was an element of a guessing game as to when that would appear, and I suggest we're about that. Now that said, we do think there will be upside, if and when we get the label, and depending on what the label says, so I don't think we have finished, but we might be finished to some extent now. The other two factors you might want to think about is that the plan with respect to duration of treatment in the adjuvant setting is a year. Now we have not talked about precise duration of treatment, but you can speculate that therefore it isn’t until fairly recently that patients were starting to come off. So for the first of the nine months everyone that was going on was additive and eventually you get to the end of the road. The other thing -- and I'm going to speculate a little bit here -- is that we did hear a lot of oncologists suggesting that they might look back at patients who have had surgery maybe six or 12 months before the date when it was first released and bring those patients back in and treat them. That would have been a one-time effect, and so there is an element of bonus of that. But they are all factored in there, and I think we believe the underlying demand will be strong. We probably will benefit from the label, because as I mentioned in the prepared remarks, the metastatic element of the market continues to grow.

Our next question comes from Mark Schoenebaum - Bear Stearns. Mark Schoenebaum - Bear Stearns: Hi, thanks very much for taking my question. Maybe I can step away from the commercial questions and since Sue is in the room maybe I can tap her expertise. Just wondering, Sue, on the U.S.-based adjuvant Avastin colon cancer trial, can you help us understand if -- and I realize this is a massive if -- but if the benefit of Avastin in the adjuvant setting is as great as the benefit was for Herceptin in its adjuvant setting, when do you think that trial might be stopped, if you take a look at the statistics? Then the corollary question is, I know that trial is enrolling both stage 2 and stage 3 patients, but if you look at prior approvals for Xolair it also enrolled stage 2s. It wasn’t actually approved for stage 2s. What do you think of the probability you could get a stage 2 label? Thanks.

So the answer to your first question is that if you look at the events in the adjuvant colon trials, depending on the magnitude of the difference and also, I must say, of the cooperative group’s ability to collect those events and to do so in a timely fashion, so that adds another factor in there that you have to take into account, the frequency of their data sweeps. We do anticipate the earliest possible date to be in 2007. So that is if the magnitude is on the order of a Herceptin-like and the events are collected very quickly. I don't want you to think that if the unblinding happens later it is only driven by the magnitude. So 2007/2008 you could see something in that magnitude. The answer to your second question is a little bit harder. If you look at the discussions we've had, for example, on the Herceptin adjuvant label, I think it's important in those, let's say lower risk patients, it depends on the number of patients who actually do enroll in the trial and how consistent the data are in lower-risk and higher-risk patients, whether or not we would expect to have that in our label. So it is the consistency of the data, the strength of the data, and just the luck of the draw of what proportion of the patients end being in that lower risk segment.

Our next question comes from Mike King - Rodman Renshaw. Mike King - Rodman Renshaw: Thanks for taking my question. Ian, I was wondering if you could give us a more detailed update on the reimbursement picture for LUCENTIS? Can you talk about what proportion of the plans you are targeting have agreed to carry it on the formulary? Do you have any experience with current payment in terms of whether docs are getting full reimbursement or if there is any kind of partial reimbursement activity taking place?

I covered quite a bit of that in the prepared remarks. We are progressing pretty well with reimbursement. Most of the carriers have agreed to cover it, if not written then verbally, and we are starting to receive, at the front end, significant claims and payment of those claims. We are not through it yet, and I think that the specialists would like a little bit more confidence in the process. But we think we have made a solid start and we are about where we expect to be at this point. Mike King - Rodman Renshaw: Can you say what that is? Is that 80% or more than 50%? Any kind of quantification?

I wouldn't want to quantify it. It is going very well though.

Our next question comes from Geoff Meacham - JP Morgan. Geoff Meacham - JP Morgan: Thanks for taking the question. My question is about the Avastin sBLA. I am just wondering what the risk that some data, such as scans, will not be recoverable? How much of the safety updates require additional follow-on data?

One of the things that we are discussing with FDA now is their expectations, not just with the number of films that need to be reviewed, but what proportion of anything missing would be acceptable to them. So we haven't finalized that with FDA. I would remind you that the magnitude, the difference in that study is very large, so any sensitivity analysis would suggest that we have a very nice move from about six months to about 13 months in times progression, so that gives me some optimism that we will be able to confirm it. But you're right in thinking about the amount of data collection we have for a relatively mature study at this point is important. In terms of the safety events, the discussions we've had with FDA are that we have a full dataset and that we go back to the sites and confirm that dataset, and not necessarily collect any additional follow-up. So it is really confirming the original database and the original dataset that we sent in to FDA that is the subject of discussion at this point.

Our next question comes from Bret Holley - CIBC World Markets. Bret Holley - CIBC World Markets: Ian, I have a question about the off-label use of Avastin, the comment that you said that it will remain a factor in the market. I guess to what extent is it going to remain a factor in under-covered patients? Could you just flesh out your thoughts there?

This is in the context of LUCENTIS. It is a little early to tell. I wanted to put that comment there to the extent, one couldn’t imagine a situation where you launched your label drug in LUCENTIS for the use of a completely unlabeled drug with almost no data that kind of disappeared overnight. Our early indications is that is not going to be the case. We haven't got detail on precisely why it was still being used. We know there are some retinal specialists who are more enthusiastic, and certainly there is an economic element in selecting the patients in some instances. But I wouldn't want to go into more detail than that based on one quarter's worth of use. But I think it is fair to say it may well be a factor for a period of time.

Our next question comes from Craig Parker - Lehman Brothers. Craig Parker - Lehman Brothers: Hopefully I can break the rules and get in a question and one-quarter maybe. The first one is just a quick clarification. Ian, when you say distribution changes didn't have a meaningful impact, are you including the change in the timing of revenue recognition? My substantive question is maybe you could help us think about the potential for LUCENTIS to have declining sales. If I think about 80% of the sales being patients switching, and that many of those patients will not receive any LUCENTIS in the subsequent quarter, it seems like there's a pretty high potential for there to be a decline in sales quarter over quarter.

The first one is straightforward. Yes, it did include the direct patient change. For me to comment precisely on a decline, it would get pretty close to being a revenue projection, which as you know we don't do. I think, though, you have caught some appropriate themes. There is going to be a number of potentially conflicting dynamics going on, having only got 8% of our patients to switch. I mean, we've got a lot more new patients to get as well. So I think what you'll see is as you see a decline in switch you'll see growth in new. You're right. It is likely that the patients we've got may well use less drug. Although there is an option to use the drug once a month as well. I'm not sure everybody will step down to a lower dose. On the other hand, I think it would be wise not to take everything that has happened in the first quarter and expect it to double and triple as we go into the second and the third quarter. So it is balancing out those factors that I think is critical in terms of thinking about it.

Our next question comes from Michael Lever - Credit Suisse. Michael Lever - Credit Suisse: Hi guys, thanks. This question is for Sue. You hinted to this in some of your comments that the FDA has some concerns over the data quality in cooperative groups. I'm just wondering if you can comment specifically in relation to the breast cancer filings? I wonder if you could comment on how that might impact your ongoing trials, your decisions to use collaborative groups in the future, and just timelines for your adjuvant trials that are using cooperative groups?

We're really in the midst of understanding the change of how FDA is looking at cooperative groups, and we're having discussions with both the FDA and the National Cancer Institute on that subject. Here's what we know now. We know now that the progression-free survival as an approvable endpoint through the cooperative groups requires a lot of diligence and discussion with the FDA before the study is carried out. That is what we know right now. The impact on our future trials is heavily dependent on whether or not we plan to submit those trials for approval, and whether or not PFS is an endpoint as compared to a survival. That is what we know. We have prioritized all of our future trials, and we will bring those to FDA for discussion, but I can't tell you the impact on the timelines yet until we have those discussions. I think in the most simple way to think about it is we will require, if we are using progression-free survival, it will require an independent radiology review. My concern is that this will decrease companies like Genentech who wish to work with cooperative groups. But we are very positive about the cooperative group. We are very positive about the thought leaders and the community participation. And so we're trying to work through this issue with FDA and NCI so that we can continue to work with those groups. Michael Lever - Credit Suisse: Would it be fair to say most of your adjuvant trials are powered in the early interim stages to look at progression-free survival, as opposed to survival as an approvable endpoint?

Most of the adjuvant trials do look at the progression-free survival. I will say that in terms of the documentation it is somewhat more straightforward for the adjuvant trials as compared to the metastatic trials, because it is the appearance of a new lesion, rather than trying to do this cross-sectional diameter with a resist criteria. Disease-free survival is more straightforward in the adjuvant and the metastatic setting.

Our next question comes from George Farmer - Wachovia. George Farmer - Wachovia: Thanks for taking my question. It has to do with Tarceva. Ian, you mentioned that second line use in lung cancer was on a decline. Can you elaborate a little bit further on that?

Not a great deal, to be honest. We try to be consistent with the way we refer to our data. In previous quarters we have had step ups of a certain magnitude, and we have called them step ups. This time we saw a step down, and I think we thought it was fair to call it a step down. You might speculate that both moves were not massive, but they were large enough to comment on. The only other factor that exists in the marketplace that we have not got a lot of data to really substantiate this is that there was the A&T data a little while ago. That was in second line setting, and that encouraged some folks to use Avastin in second line lung. We have no other wider explanation. I wouldn't see that move as a fairly dramatic move. George Farmer - Wachovia: But should we think about that going forward? Is there going to be increased second line penetration, do you feel, as say more A&T data is accumulated or is there a potential that there's competition from the cytotoxics that are on the market?

I think what you see with most of our cancer drugs is you see share evolution fairly rapidly after data with a label, and then you see some consistency until there's a major change in the market. We are a ways away from any significant new Tarceva data, as Sue commented in her prepared remarks. I don't see anything dramatically new from any of our competitors either. I would speculate that the dynamic has not changed dramatically in the short to medium term.

Our next question comes from Jason Kantor - RBC Capital Markets. Jason Kantor - RBC Capital Markets: Hi, thanks. Could you confirm that the Avastin sales number reflects a decrease in end-user demand? Could you quantify how much of the use was coming from these other indications like renal cell and pancreatic, and if there is more downside from those indications?

No, I can't confirm because I didn't say that. I said actually that overall use has gone up and sales did go up as well. We obviously had a number of different dynamics in there. We had some sales which were accruing from off-label pancreatic use where we have the negative study, and some sales in renal cancer where the drugs that were approved have taken some of those sales away; although to be fair, they have also expanded the market. So on balance, use went up. We didn't have a lot of use in those other areas, so I wouldn't expect that part of dynamic to go on very much either.

Our next question comes from Geoffrey Porges - Sanford Bernstein. Geoffrey Porges - Sanford Bernstein: Thanks very much. A question for Sue. You have seen a lot of different studies now for Avastin in both Phase II and Phase III. Could you comment on whether you believe there is a dose response to the drug? Then whether we should anticipate over time that you might move to a 10 mg per kgs dose in front line colorectal cancer and whether you would explore other even higher doses in other indications in the future?

Having looked at all the pre-clinical data and clinical data a few months back that one of our clinical scientists put together, I would say that my personal belief is that more is better. In terms of efficacy we always try and carefully balance that with the safety. We do know that at double the dose that is an average of 5 mg per kg per week, we did run into dose-limiting toxicity in our early studies in breast cancer. So I do think there's a limit, but I believe given the data we have up to now that the 5 mg per kg per week, in general, has translated into better benefit in both pre-clinical and clinical trials, and also treatment to progression and a suggestion that perhaps even beyond progression may be beneficial to patients; but more speculation than data, I have to say. But in our trials we have consistently tried to use the dose that is more the 5 mg per kg per week if tolerated because of that data-driven aspect. We will see the information from the two Roche trials I mentioned in breast cancer and lung cancer. Before we make a lot of additional plans to study any higher doses in any new indications, I would say that we will look very carefully at those trials to see if there's any additional dose response data to the positive or the negative.

Our next question comes from May-Kin Ho - Goldman Sachs. May-Kin Ho - Goldman Sachs: Hi, Ian. Now that we are close to the approval of Avastin for lung cancer, and that the approval might be now separated from the first-line breast cancer approval, can we assume that there is not going to be any major change in pricing for Avastin?

We have said that when we get the approvals in both lung and breast that we would talk about how we would deal with the fact that they are both at a higher dose, and for good reasons, as Sue just talked about. In terms of addressing that, as you know, we have a co-pay program which is growing. We have our Access to Care Program. We have said that if we are going to take any other steps we would announce those steps at the granting of the label. So if we were to be doing so we would announce that tomorrow as we receive the label. Sorry if I can't be more forthcoming than that today.

Our next question comes from Thomas Wei - Piper Jaffray. Thomas Wei - Piper Jaffray: Thanks very much. I had a question on LUCENTIS. Do you have a guess as to what percentage of the pool of the pre-existing AMD patients you may have already penetrated in the third quarter? In particular, if you try to account for any of those patients who are too far gone in terms of loss of vision to really be treated with LUCENTIS? Is there still a lot of patients left over there to be treated or have you largely penetrated that opportunity already?

You have two questions there. The first in terms of penetration to the existing market, I didn't quantify that on the call. It is a little early to do it. I think it is fair to say there is some way to go yet within that penetration. In terms of patients whose disease have progressed, and they typically refer to it as significant scarring, then even using LUCENTIS there's little evidence that you can correct any improvements. I don't think we see that there's going to be an expansion of the market by patients being called back in with significantly progressed disease.

Our final question comes from Gene Mack - HSBC Securities. Gene Mack - HSBC Securities: Sue, you alluded to potentially getting additional updates of timing on some of the cooperative group studies. I am just wondering, are you going through each study on a case-by-case basis with the FDA? Are there some things you can do proactively, at least away from the progression-free survival endpoint, but maybe just in terms of data collection, are there things you can do proactively to minimize any kind of adverse impact to timing?

Definitely we already understand some of the comments that FDA has made in some of their requests, and so we have already discussed with both the National Cancer Institute and the cooperative groups these changing standards. So that is happening already on a bigger picture level. It is really with regards to having end of Phase II meetings and bringing that end of Phase II discussion closer to what we typically see with industry sponsored trials that we have to request those meetings, set those up, and so those meetings in and of themselves impact the timeline. But we are being as proactive as possible, given our current understanding of the expectations, to move all of those studies more in line with FDA expectations.

Ms. Littrell, do you have any closing comments?

Yes. I want to thank you all very much for joining us today. Sue Morris and I and Diane Schritck will all be in our offices in the next few minutes, so feel free to give us a call if you have further questions. Thank you very much.

Ladies and gentlemen, this concludes the Genentech Q3 2006 earnings conference call.