Forty-six children with steroid-responsive nephrotic syndrome were randomly allocated to receive two different prednisolone regimens for initial therapy. Twenty-nine children (group 1) received an intermittent regimen (60 mg/m2/day for 4 weeks, followed by 40/mg/m2/day on 3 days a week for 4 weeks); 17 children (group 2) had a long-term regimen (60 mg/m2/day for 4 weeks, followed by the same dose on alternate days for 4 weeks and the doses tapered by 10 mg/m2, given on alternate days every 4 weeks for 5 months). There was no difference between the two groups in the regimen used to treat relapses, steroid responsiveness, number of patients with relapses, and frequency of toxic reactions to steroids. However, the number of patients with a relapse within 6 months after initial therapy and the number of those with frequent relapses or steroid dependence were significantly higher in group 1 than in group 2 (P less than 0.05 for both). The data indicate that the long-term tapering regimen appears to be both safe and preferable to the intermittent regimen for initial therapy in children with idiopathic nephrotic syndrome.