Because of its potential for chronic nephrotoxicity, the long-term use of cyclosporine A (CsA) as treatment for nephrotic syndrome (NS) is controversial. The clinical outcome of patients with NS treated with CsA is unclear. We retrospectively evaluated 35 children with idiopathic NS, 24 with steroid-dependent NS (SDNS), and 11 with steroid-resistant NS (SRNS), who received CsA therapy for more than 12 months (median, 23 months) at the dosage maintaining 50 to 120 ng/mL in trough level. For SDNS patients, CsA was added to prednisolone after complete remission was achieved. For SRNS patients, CsA was used in combination with alternate-day prednisolone. Initial renal histology showed minimal changes (MC) in 28 patients (including all of the patients with SDNS) and focal segmental glomerulosclerosis (FSGS) in seven patients. The patients were followed up for 2 to 10.5 years (median, 6.5 years) after the termination of the CsA therapy. In SDNS patients, the relapse rate, dosage of prednisolone, standard deviation score for height, and body mass index significantly improved during CsA treatment. The follow-up study showed the proportion of SDNS decreased to 13 of 24 (54%) patients. In SRNS patients, CsA therapy induced remission in 8 of 11 patients (73%) (complete remission in seven and incomplete remission in one). Six of 11 patients (55%) then became steroid sensitive. Post-therapy biopsies, performed in 13 patients (10 with SDNS and three with SRNS), showed mild stripped interstitial fibrosis in two SDNS patients (15%). Long-term CsA therapy in moderate doses was effective to the patients with SDNS and SRNS and low in incidence of nephrotoxicity. The long-term use of CsA appears to result in a decrease in the proportion of SDNS and acquisition of subsequent steroid responsiveness in SRNS.