Steroid-sensitive nephrotic syndrome often follows a relapsing course with a substantial number of patients requiring cytotoxic therapy with cyclophosphamide (CP). However, the long-term success of CP treatment is difficult to predict. We retrospectively evaluated 106 patients after CP to identify parameters associated with sustained remission. The overall rate of cumulative sustained remission was 24% after 10 years. No gender difference was found. Several factors were significantly correlated with the rate of sustained remission: age at CP therapy (remission 34% versus 9% in children older or younger than 5.5 years, P<0.01), frequently relapsing versus steroid-dependent status (54% versus 17%, P<0.05), leukopenia under CP treatment (44% in children with leukopenia versus 19% in children without leukopenia, P<0.05), and a cumulative dosage per body surface area (BSA) of more or less than 5,040 mg/m(2) (45% versus 11%, P<0.01). In contrast, the cumulative dosage per kilogram body weight had no significant influence on long-term remission (23% in children with >168 mg/kg versus 26% in children with <168 mg/kg, P>0.05). The current concept of CP treatment of steroid-sensitive nephrotic syndrome is less effective in preschool children. CP therapy should be re-evaluated on a BSA-adjusted regimen.