3511 Background: The 3-year results of NSABP protocol C-03 published in 1993 (Wolmark, et al J Clin Oncol 11:1879–87, 1993) indicated significant improvement in DFS and OS for patients (pts) with stage II or III colon cancer treated with postoperative FL compared to pts treated with lomustine (MeCCNU), vincristine, and 5-FU (MOF). The present analysis reports long-term results. Methods: Pts with stage II or III (Dukes’ B or C) colon cancer who had undergone a potentially curative surgical resection were randomized postoperatively to receive adjuvant chemotherapy with either MOF or FL (given on the “Roswell Park schedule”) for approximately 1 year. Results: 1081 pts (1044 eligible-96.6%) were randomized into C-03 from 1987–89. Of eligible pts 986 (94.4%) had complete 10-year data (event or complete follow-up) and 1022 (97.9%) had complete 5-year data. The 10-year Kaplan-Meier estimates, relative risks from a stratified Cox model, and stratified logrank p values (stratified for gender, # positive nodes [0, 1–4, 5+], and tumor location [right colon, non-right], as specified in the protocol) for DFS and OS are shown below. Tests for interactions of the stratification variables (gender, # positive nodes, and tumor location) with treatment were not significant (p>0.16). Conclusions: The 10-year results of NSABP C-03 confirm and extend the previously published analysis, and indicate a highly significant and clinically relevant improvement in DFS and OS for pts with stage II or III colon cancer treated with FL compared to MOF. These data serve as a standard for evaluation of new adjuvant regimens. No significant financial relationships to disclose.