Despite the accumulation of positive data, the role of azathioprine (AZA) in the maintenance of remission of ulcerative colitis is still controversial. We looked at the follow-up of the ulcerative colitis patients who, after responding to either steroids or cyclosporin (CsA), received AZA at our referral center for over a decade. The 39 patients (29 m/10f) were treated between 1991 and 2007. Twenty-five of them had responded to CsA, the remaining 14 to corticosteroids. AZA was usually overlapped with either of the two agents at the initial dose of 2mg/kg/day. The definitions of remission, relapse, and AZA toxicity followed commonly agreed criteria. The median duration of the AZA treatment was 14 months (<1-201). Fifty-two percent and 14%, respectively, of the CsA and the steroid responders needed surgery (overall rate=38%). The figures were 32 and 15 at the first year. The majority of the patients had 1-2 relapses often in connection with withdrawal of AZA; only 3 of these relapsers needed hospitalization. AZA caused toxicity in 16/39 (41%) patients, requiring withdrawal in 23% of the cases; leukopenia (17%) and hepatitis/cholestasis (10%) ranked first and second for frequency. All of the patients in whom AZA was stopped (or reduced) relapsed. In conclusion, the 1-year colectomy rates compare favorably with the figures reported by the literature. By contrast, the toxicity rates were higher than expected. Failure to genotype or to use escalating AZA doses can only be hypothesized as causes.