In a double-blind controlled trial of 194 patients with clinically definite active multiple sclerosis, 98 were randomized to treatment with cyclosporine (CyA, 5 mg/kg/day), and 96 to treatment with azathioprine (Aza, 2.5 mg/kg/day). Eighty-five patients in the CyA group and 82 in the Aza group completed a treatment period of 24 to 32 months in accordance with the study protocol. No significant differences could be detected between the two treatment groups at the end of the trial. Assessment was done by serial quantitative neurological examinations and Kurtzke's Expanded Disability Status Scale. Frequency of relapse and patient self-evaluation also failed to show significant differences. Overall deterioration observed in both groups during the trial was only minor. The incidence of side effects in the CyA group was more than two times that in the Aza group. We conclude that CyA as a single agent cannot be the drug of final choice in long-term immunosuppressive treatment of relapsing-remitting and relapsing-progressive multiple sclerosis.