The purpose of this study was to evaluate the efficacy and safety of alternate-day prednisone therapy in treating patients with mild-to-moderate cystic fibrosis during a 4-year period. In 15 North American cystic fibrosis centers, we screened 320 patients and enrolled 285 patients from April 1986 to December 1987. Patients were randomly assigned to take prednisone, 1 mg/kg per dose or 2 mg/kg per dose, or a matching placebo given on alternate days. Lung function, clinical status, hospitalizations, growth, and steroid side effects were monitored. During the first 24 months the percentage of the predicted forced vital capacity was greater in the 1 mg/kg group (p < 0.0001) and the 2 mg/kg group (p < 0.01) when each was compared with placebo. Patients in the 1 mg/kg group had a higher percentage of predicted forced vital capacity than placebo patients during the entire 48 months (p < 0.0025), but only in the group of patients who were colonized with Pseudomonas aeruginosa at baseline. For 48 months, the 1 mg/kg group had a higher percentage of predicted forced expiratory volume in 1 second than patients given placebo (p < 0.02). The prednisone-treated groups had a reduction in serum IgG concentrations (1 mg/kg vs placebo, p < 0.007; 2 mg/kg vs placebo, p < 0.003). From 6 months onward, height z scores fell in the 2 mg/kg group compared with those given placebo (p < 0.001). For the 1 mg/kg group, height z scores were lower from 24 months. An excess of abnormalities in glucose metabolism was seen in the 2 mg/kg group compared with the placebo group (p < 0.005). Our findings suggest a role for alternate-day prednisone therapy at a dose of 1 mg/kg in patients with mild to moderate cystic fibrosis. The benefit of improved lung function appears to outweigh the potential for adverse effects when the treatment period is less than 24 months.