A series of 274 attacks (256 of them recurring) of ulcerative colitis in 158 subjects, observed between January 1974 and December 1990 in Bratislava (48 degrees N, 17 degrees E), Slovakia, has been processed by R.A. Fisher's periodogram and F. Halberg's cosinor procedure. In spite of the long-term administration of sulfasalazin and/or corticosteroids, an unequivocal statistically significant (alpha = 0.01) annual (P = 0.0003) rhythmicity was identified. The amplitudes of both cycles were very similar, approximately 25% of the MESOR. The first estimated peak occurred in the second half of March and the first half of April, with an estimated increase of more than 50% over the average monthly incidence, resulting from coincident acrophases of both components. The second semiannual peak occurred at the end of September and early October. It is concluded that the long-term chemotherapy, applied in the majority of cases, failed to substantially influence the known circannual variations of the disease attacks. The issue has to be taken into account in planning prevention measures, treatment and its evaluation, as well as in organizing the regimen of management in endoscopic centers.