Evidence points to possible cohort effects in inflammatory bowel disease, the possible role of perinatal infection as a risk factor for inflammatory bowel disease, and the occurrence of clusters of Crohn's disease. This evidence suggests the value of searching for birth date clustering among cases of inflammatory bowel disease. The authors looked for clustering by birth date and maternal residence at birth in a population-based series of 845 Crohn's disease patients and 1,330 ulcerative colitis patients born from 1924 through 1957 and diagnosed in the Uppsala Health Care Region, Sweden, until the end of 1983. Over this period, 43% of persons with Crohn's disease had been born within 6 days of another case, compared with 36% of controls simulated to account for monthly variation in births (p = 0.0002). The number of pairs of inflammatory bowel disease cases whose births occurred in the same county (close in space) and whose birth dates were also close in time was statistically significantly greater than expected for most birth dates 23-57 days apart. Results after 1944, when ascertainment was more complete, generally corroborate these findings and suggest some seasonality in the birth dates of ulcerative colitis cases. Results from the entire study period and after 1944 thus provide evidence for clustering by birth (including seasonality) among Crohn's disease cases and also, to a lesser extent, among ulcerative colitis cases.