OBJECTIVE Immunosuppression with methotrexate may be useful in the treatment of Crohn's disease. We tested the efficacy of methotrexate in refractory Crohn's disease in a randomized, controlled trial. METHODS Randomized, double-blind placebo-controlled trial of methotrexate in 33 patients with steroid-dependent Crohn's disease, 33% of whom had previously failed therapy with 6-mercaptopurine. Patients were given placebo or oral methotrexate 15 mg/week, or adjusted up to 22.5 mg/week, for up to 1 year or until treatment failure. Outcome was assessed by reduction in prednisone dosage, Crohn's Disease Activity Index, hospital admission, and laboratory parameters. RESULTS Four patients were dropped from the study for non-compliance and one because of intercurrent illness, and 28 patients could be evaluated. Fewer methotrexate-treated patients (6/13 or 46%) had flares of Crohn's disease as compared to placebo-treated patients (12/15 or 80%), but this did not achieve statistical significance (p<0.1). There was a non-significant trend toward an increased number of significant side effects in the methotrexate-treated patients (3/13 or 23%) as compared to the placebo-treated patients (0/15 or 0%) (p<0.2). Laboratory indices of inflammation did not differ between the two groups. CONCLUSIONS The methotrexate-treated group showed a trend toward fewer Crohn's disease flares, balanced by an increased number of significant side effects.