This study sought to replicate and expand findings reported by the National Surgical Adjuvant Breast and Bowel Project (NSABP) on prognostic factors in resectable colon cancer. Mantel-Haenszel tests and the Cox model were used to analyze prognostic significance and effect of primary disease symptoms and tumor location in 572 patients from the Gastrointestinal Tumor Study Group (GITSG), with resected Dukes' B2 and C colon cancer. Tumor location (left, right, and rectosigmoid/sigmoid) was of low prognostic importance (P greater than 0.10), and did not effect survival or disease-free survival (P greater than 0.10). Obstruction was an important indicator of prognosis, independent of Dukes' stage (P = 0.03). Bowel perforation is associated with poor prognosis in disease-free survival (P = 0.001). Rectal bleeding had a positive impact on survival (P = 0.08). Thus, obstruction, perforation, and rectal bleeding (but not location) are found to be prognostic factors in patients with Dukes' B2 or C colon cancer.