Experimental colitis was induced in guinea pigs by administration of 5% degraded carrageenan for 5 days. The prophylactic effect of a slow-release preparation of 5-aminosalicylic acid (5-ASA; 13 mg/100 g/day) was compared with approximately equimolar amounts of salazosulphapyridine (SASP; 26 mg/100 g/day) and placebo. Treatment was started 2 days before initiation of carrageenan administration. The drugs were administered through a chronic gastric fistula. At the end of the study concentrations of 5-ASA and acetylated 5-ASA (Ac-5-ASA) in cecal contents and in plasma were determined. In the placebo group, all guinea pigs developed many small punctiform ulcerations in the cecum (median, 30/cm2). In the 5-ASA group no protective effect was demonstrated, since the number of ulcerations was 37/cm2. The difference is not statistically significant. However, the SASP group presented significantly fewer ulcerations (4/cm2). The concentrations of 5-ASA and/or its acetylated metabolite were several times higher in the cecum content and twice as high in plasma in the SASP group, indicating a difference in the absorption patterns of 5-ASA and the two drugs. These results and the etiological difference between the human ulcerative colitis and the carrageenan model may account for the lack of prophylactic effect of the slow-release 5-ASA in this experiment.