The response to 5-aminosalicylic acid (5-ASA) in active Crohn's disease was studied in comparison to methylprednisolone in a 24 week randomized double-blind multicentre study. Sixty-two patients were included in the analysis. Thirty were treated with 500 mg 5-ASA qid and 32 with methylprednisolone (starting dose 48 mg for one week, then reduced weekly to 32, 24, 20, 16 and 12 mg with maintenance at 8 mg/day for the remaining 18 weeks). Mean age, earlier surgical intervention, localization of Crohn's disease and extraintestinal manifestations were not different in both groups. The Crohn's disease activity index (CDAI) and the van Hees index were not significantly different in both treatment groups at the entrance examination (median CDAI 232 in the 5-ASA group and 220 in the methylprednisolone group). According to the protocol, treatment was stopped due to insufficient efficacy in 73% of the patients receiving 5-ASA and in 34% of the patients receiving methylprednisolone (x2 test P=0.0019). The area under the curve for the CDAl was significantly greater in 5-ASA (median 170) than in methylprednisolone (P≤0.007) (68). Eleven per cent of patients taking 5-ASA and 26% of patients taking methylprednisolone presented relevant side effects to treatment (not significant). It is concluded from these data that 5-ASA at the dose used in this study is not efficient in the treatment of active Crohn's disease. Considering recent studies in ulcerative colitis, a trial using a higher dose is indicated.