The epidemiology of ulcerative colitis (UC) in Stockholm County 1955-79 was investigated. There were 1274 cases. The proportion of patients with proctitis, left-sided and total extent of disease remained constant over the study period. The incidence increased over the first 20 years followed by a plateau. The peak incidence in relation to age increased slightly but remained in the third and fourth decade through the study period. The incidence in men over 40 years of age increased markedly towards the end of the study. There were 109 deaths at follow-up (Dec. 1981) among those having the UC diagnosis between 1955-1979. Twentysix out of 41 patients who died due to UC did so postoperatively. The mortality pattern among those 68 patients who had causes of death unrelated to UC was similar to the expected. There was a decrease in cumulative survival probability compared with the expected, in particular in those with total colitis. This was also seen when only deaths unrelated to UC were included suggesting an increased sensitivity to the ordinary disease spectrum in UC patients. Between 1945-1979 there were 1339 patients who had UC diagnosed. Twentyfive of those had developed colon cancer (24 with total colitis) at follow-up (Dec. 1981). Eighteen were dead at follow-up, the survival time being in direct relation to the Dukes' grading at cancer diagnosis irrespective of age. The cumulative cancer risk at 25 years duration (total colitis patients only) was 13% (SD 4%) compared to the expected 1,9%. In a selected group of 71 patients followed for cancer surveillance during the period 1974-82 nine patients developed at least low grade dysplasia including one Dukes' A carcinoma. The dysplasia or cancer leading to operation was found above the rectum in four of five operated patients, all having total colitis with a duration ranging between 25-44 years.