Cytotoxic agents have long been used in Behçet's syndrome, especially for eye involvement, but their effectiveness has been uncertain. We conducted a two-year randomized, placebo-controlled, double-blind trial of azathioprine (2.5 mg per kilogram of body weight per day) in Turkish men with Behçet's syndrome without eye disease (group 1; n = 25) or with eye disease (group 2; n = 48). Corticosteroid treatment remained available to all the patients. All six patients withdrawn from the study because of severe eye disease were receiving placebo (P less than 0.001). Azathioprine was superior to placebo in the prevention of new eye disease in group 1 (1 vs. 8 patients; P less than 0.01) and in group 2 among the 14 patients who at entry had disease in only one eye (P less than 0.001). There were fewer episodes of hypopyon uveitis (1 vs. 15; P less than 0.001) among the group 2 patients who took azathioprine. The patients taking azathioprine also had less frequent oral ulcers, genital ulcers, and arthritis. There were no serious side effects attributable to azathioprine. We conclude that azathioprine is effective in controlling the progression of Behçet's syndrome, especially its most serious manifestation, eye disease.