6037 Background: Historical survival evaluation of clinical trials for advanced NSCLC suggests that median survival improved by 0.4 months over a 22 year period. Survival in clinical trails is evaluated from treatment initiation and not from diagnosis. If treatment is initiated earlier over time, this may create a lead-time effect on survival analysis and thus may explain the apparent discrepancy between trial findings and outcomes in the community. Objective: Investigate if, over time, there has been a shortening of the time between the diagnosis of NSCLC and the initiation of chemotherapy. Methods: We evaluated patients with NSCLC stages IIIB and IV between 01/01/1994 and 12/31/1999. Date of diagnosis was obtained from Surveillance, Epidemiology, and End Results (SEER). Treatment data was obtained from Medicare claims and is restricted to patients above age 65. Differences were compared using analysis of variance (ANOVA). Results: A total of 11,995 patients were identified. Patient’s characteristics: median age 74 years; 57% of patients were males; racial distribution: white (82.4%), black (9.4%), Asian (3.2%), Hispanics (1.2%), others (3.8%); stage distribution: IIIB (38%), IV (62%). Chemotherapy was given to 3,377 (28%) patients. There was an increase in chemotherapy utilization over time: 10% in 1994 to 35% in 1999. Over time there was a shortening from diagnosis to chemotherapy initiation from 1.6 months in 1994 to 1.3 months in 1999 (p = 0.0004). Among those receiving chemotherapy, median survival from diagnosis did not change over this time period. Black patients had a longer time between diagnosis and treatment initiation: 1.7 months vs. 1.4 months (p < 0.01). Conclusions: Over time, there has been a shortening of the time between diagnosis and chemotherapy initiation. This difference, although small, seems to account for much of the apparent historical survival benefit observed. This finding puts additional limitation on the historical comparisons of survival between trials. Racial differences may help explain, in part, observed differences in outcome. No significant financial relationships to disclose.