To compare the efficacy (induction of remission) and safety of cyclosporine (CsA) with those of supportive therapy in patients with steroid-resistant idiopathic nephrotic syndrome (INS), we organized an open, prospective, randomized, multicentric, controlled study for parallel groups, stratified for adults and children. Forty-five patients with steroid-resistant INS were randomly assigned to supportive therapy or CsA (5 mg/kg/day for adults, 6 mg/kg/day for children) for six months, then tapered off by 25% every two months until complete discontinuation. Four patients were lost to follow-up. During the first year 13/22 CsA-treated patients versus three of 19 controls attained remission of the nephrotic syndrome (P < 0.001). A symptom score was assessed at time 0 and at six months. The mean score significantly decreased in the CsA group (P < 0.001), but remained unchanged in the controls. At month 6 the mean urinary protein excretion, the mean serum proteins and plasma cholesterol had significantly improved in the CsA group but were not changed in the controls. There were no significant differences in serum creatinine and creatinine clearance between treatments (interaction time* treatments, P = 0.089 and P = 0.935, respectively) at month 6 versus basal. The CsA-related side-effects were mild; no significant difference in blood pressure between the two groups was seen at any time. This study shows that CsA can bring about remission in some 60% of patients with steroid-resistant INS. In patients with normal renal function and without severe hypertension, CsA at the therapeutic scheme adopted did not produce severe renal or extrarenal toxicity.