Forty-nine patients with membranous nephropathy (MN) and nephrotic syndrome (NS) were randomly allocated to supportive or specific therapy. The latter consisted of steroids or chlorambucil given in alternate months for a cumulative period of six months. Three patients in the experimental group were dropped from the study because of therapy related side-effects. At the end of follow-up there were significantly more patients in complete or partial remission in the experimental group than in the controls. The mean serum creatinine did not change in treated patients, but it significantly increased in controls.