The epidemiological understanding of inflammatory bowel disease has been limited by the referral bases of most inflammatory bowel disease studies. The Colitis-Ileitis Study Group of Rochester, N.Y., developed a community-wide, computerized cumulative registry of all inflammatory bowel disease patients hospitalized at the 8 community hospitals for 1973-86. Clinical data were abstracted from each of the 1651 identified hospital charts. All of these patients resided in Monroe County (city and suburbs) and the 5 contiguous counties (Genesee/Finger Lakes Region, population 1,030,640). Of the 1651 hospital patients identified in the study, 1358 resided in Monroe County (Rochester and its immediate suburbs, population 702,238). Incidence, defined as time of onset of symptoms of inflammatory bowel disease, rose from baseline rates in the 1930s to peak in 1980 (Crohn's disease = 50.29/10(5) per decade, ulcerative colitis = 35.12/10(5) per decade) and declined through 1986. For Crohn's disease, the age-specific incidence rates peaked in the 20-29-yr-old group in each of the 5 decades studied. Ulcerative colitis seems to occur at all ages and may have a bimodal distribution. There was a period effect, with the 1970s having the highest incidence of Crohn's disease and ulcerative colitis for each age group. However, the age-specific incidence rate for Crohn's disease showed a 40% decrease in the 1980s compared with the 1970s in the 10-39-yr-old group (p less than 0.001). The age-specific incidence rate for ulcerative colitis showed a 50% decrease in the 1980s compared with the 1970s in the 10-49-yr-old group (p less than 0.001).