The Medical Record departments of the five teaching hospitals in Edmonton, plus the 37 community hospitals in the eight census districts of the northern half of the province of Alberta, Canada, were contacted, and a search was made of all patients with a discharge diagnosis of Crohn's disease or ulcerative colitis. Also, the patient records of all Edmonton gastroenterologists were reviewed to discover patients with Crohn's disease or ulcerative colitis who had never been hospitalized within these census areas. From January 1, 1977, to December 31, 1981 (which was the prevalence date), the population was 1,295,360. Of the 2,419 patients with inflammatory bowel disease, 48.5% had definite Crohn's disease and 33% had definite ulcerative colitis. There were 1,716 (70.9%) patients analyzed in this study. The factorial analysis of disease prevalence per 10(5) population revealed that significant differences were found for location of residence, sex, and age. The prevalence of Crohn's disease was higher in urban than in rural areas and in females than in males, whereas the prevalence of ulcerative colitis was unaffected by these variables. The peak prevalence of Crohn's disease was below age 29 in males and females, and the prevalence in young women at this age was approximately twice that in males. The highest prevalence of Crohn's disease was in urban females aged 20-39 (greater than 234 cases/10(5) population), with similar prevalence rates in urban males and rural females, and with the lowest prevalence rates in rural males. The incidence of Crohn's disease was greater than for ulcerative colitis, began to increase in about 1965, and reached a plateau in the late 1970s. In conclusion, the demonstration of an age, location of residence, or effect of sex on the prevalence of inflammatory bowel disease requires multiple factorial analyses. When the sample is extrapolated to the total diseased population of the region, a prevalence value of 330/10(5) was derived for young female urban individuals residing in this northern area.