In a review of 1248 cases of ulcerative colitis seen at the Cleveland Clinic that were followed up to 1984 (mean, 14.4 years), 82 patients (6.5%) were subsequently found to have colorectal cancer and 48 (3.8%) had extracolonic malignancy, 6 of them with associated colorectal cancer. Most patients with colorectal cancer were men (2:1), and had extensive (90%) and long-lasting colitis (10 years or more in 93% of cases; mean 18 years). Colitis was inactive before the diagnosis of cancer in 48%. Acute onset of the first attack was rare (7%), and the disease had a remittent course in 92%. The mean age at diagnosis of cancer was 43 years. The cumulative risk of colorectal cancer was significantly higher in patients with extensive colitis than in those with left-sided disease (P less than 0.0001: 11.9% vs. 1.8% at 20 years and 25.3% vs. 3.7% at 30 years). When comparing mean duration of disease, left-sided colitis (22 years) did not differ significantly from extensive disease. The tumor was multifocal in 13.5%, proximal to the splenic flexure in 44%, and poorly differentiated in 34% of the cases. The diagnosis was suspected clinically in 64% of cases. The prognosis of colorectal cancer in patients with ulcerative colitis appears to be similar to that in the general population. The cumulative 5-year survival rate was 54%. This study supports the concept that surveillance colonoscopy should be started after 8 to 10 years of extensive colitis and after 15 years of left-sided colitis. Among those with extracolonic malignancy, the incidence of bile duct carcinoma, leukemia, bone tumors, and endometrial cancer was significantly greater than expected (P less than 0.01), whereas that of lung cancer was significantly lower than expected (P less than 0.01).