A diagnostic and surveillance program using colonscopy in patients with colorectal cancer was established at North Carolina Memorial Hospital. The records of all patients who had preoperative or postoperative colonoscopic examination between 1976 and 1979 were reviewed. Fifty-five patients had colonscopic examination preoperatively. No additional disease was found in 39. In 15 patients, unsuspected additional disease was detected, and one patient had a suspected polyp ruled out by colonoscopic examination. One of these patients was found to have a synchronous primary cancer, not demonstrated by barium enema. Surgical treatment was modified in nine (16%) of these 55 patients by the preoperative colonoscopic findings. Sixty patients had colonoscopy six months to six years postoperatively. No additional disease was found in 47. Adenomatous polyps were found in eight. Two patients had recurrent cancer proved by colonoscopy, and three had a second primary cancer detected only by colonoscopy. Treatment was directly influenced by colonoscopy in eight (13.3%) of these 60 patients. These studies had a favorable cost/benefit ratio in patients with colorectal cancer and support a program of preoperative colonoscopy in patients with colorectal cancer and reexamination within two to three years after operation.