Fifty-four children with steroid-responsive nephrotic syndrome relapsing within 6 months of their initial response were randomly allocated to receive two different regimens of prednisone therapy. The test regimen differed from the standard prednisone relapse regimen used by the International Study of Kidney Disease in Children in that the total dosage was about double, and the duration of daily therapy (8 weeks vs. a mean of 12 days) as well as the total duration of treatment (8 weeks vs. a mean of approximately 6 weeks) was longer. The proportion of patients relapsing during treatment was significantly smaller (8% vs. 40%) and the length of remission following treatment was significantly longer (3.27 vs. 1.48 months) in the test group. All patients in both groups relapsed by 8 months. During a period of approximately 6 months after this relapse, neither the frequency of relapses nor the mean number of days of proteinuria differed significantly. Opinions of participants in this multicenter trial varied concerning whether these statistically significant differences clinically justified exposing patients to the more intensive treatment regimen. However, all agreed that neither form of treatment was satisfactory in terms of preventing subsequent relapses.