Eighteen of 60 patients with persistent moderate proteinuria between 1.0 and 2.0 g/day and a long-term follow-up of more than 4 years received steroid therapy for a mean period of 18 months. Fifteen of these 18 patients maintained their initial creatinine clearance (C<sub>cr</sub>) of 70 ml/min or more. The remaining 42 received antithrombocyte drugs and/or nonsteroidal anti-inflammatory drugs, with 31 of them keeping their initial C<sub>cr</sub> values of 70 ml/min or more. All 14 cases with initial C<sub>cr</sub> values of less than 70 ml/min in both the steroid and nonsteroid groups followed progressive courses, with 12 ending up in hemodialysis. Of the nonsteroid, preserved group of 31 cases, 12 followed a stable course, 10 a progressive course, and the other 9 went into end-stage renal failure necessitating hemodialysis. There were no differences in initial clinical features among these three subgroups. But histological changes were milder in the stable subgroup than in the progressive and hemodialysis subgroups. Further, the total score of eight histological parameters was 6 or less in all but 1 of the cases of the stable subgroup, but 8 or higher in the other two subgroups. Therefore, in the progressive cases alone with initial C<sub>Cr</sub> values of 70 ml/min or more and high total scores of 7 or more, the long-term clinical courses were compared between steroid and nonsteroid groups (10 and 20 cases, respectively). In the steroid group proteinuria decreased significantly and renal function was preserved well as compared with that in the nonsteroid group. There were significant differences in the degrees of proteinuria in the 1st year (p < 0.05) and in the C<sub>cr</sub> values 3 years after initiation of therapy (p < 0.05) between the two groups. These results indicate that appropriate steroid therapy is beneficial even in progressive cases with moderate proteinuria and moderate to severe histological alterations, if only initial renal function with C<sub>cr</sub> values of 70 ml/min or more is preserved.