7573 Background: Cutaneous rash often develops in erlotinib-treated patients (pts) and is thought to be associated with improved response and survival. This analysis focuses on incidence, severity, and time course of skin rash secondary to erlotinib, as well as the correlation between rash development and overall survival (OS) in pts in the NCIC CTG BR.21 (NCT00036647) trial. Methods: Pts with stage IIIB or IV non–small cell lung cancer (NSCLC) who had failed first- or second-line chemotherapy were randomized 2:1 to erlotinib or placebo. Cutaneous toxicity was graded per NCI CTC v2.0 and managed at the discretion of the treating investigator. This retrospective analysis used a landmark approach, dividing pts into rash and no-rash groups, by the appearance of rash by week 10. Results: A total of 366 of 488 erlotinib-treated pts (75%) and 40 of 243 placebo-treated pts (16%) developed rash at any time point. Of the former, 75% of rash episodes occurred by week 2. The incidence of grades 1 and 2 rash peaked at week 3; grade 3 rash peaked at week 5. The erlotinib dose was reduced in 48 pts (10%) due to rash. In most erlotinib-treated pts with rash, severity decreased or plateaued over time (Table). Some 311 erlotinib-treated pts (rash group) developed rash by week 10. Median OS in the erlotinib-treated rash (n=311) and no-rash (n=65) groups were 37.4 and 11.1 weeks, respectively (hazard ratio=0.51 [95% confidence interval, 0.38–0.68]; P<.0001). Baseline factors significantly associated with prolonged OS in the rash population included race (Asian vs white), number of affected organs (<3 vs ≥3), and smoking status (never vs ever). Conclusions: For most pts, rash secondary to erlotinib presented within the first 2 weeks, peaked during weeks 3–5, decreased thereafter, and did not necessitate dose reduction, thus supporting SATURN results (Perez-Soler, ASCO 2011, abst. 7610). Development of rash by week 10 of erlotinib therapy was associated with significantly improved OS. [Table: see text]