In a double-blind controlled trial 43 patients with relapsing-remitting multiple sclerosis were treated either with anti-lymphocyte globulin, prednisolone, and azathioprine, or with placebo preparations. Treatment began with a combination of the three medicaments but after 1 month was continued for another 14 months with azathioprine (3 mg/kg dialy) only. There was a marginally beneficial effect of immunosuppression on the overall relapse rate and clinical progression. However, there were significant effects on in-vitro lymphocyte function and in the visual evoked potentials in favour of the group receiving suppressive treatment. Placebo-treated patients of the HLA A3 tissue type had significantly more relapses than placebo-treated patients who were not of type HLA A3. Nevertheless, HLA-A3-positive patients treated with immunosuppression had significantly fewer relapses than A3-positive placebo-treated patients.