Although therapy with intravenous (IV) rituximab and tacrolimus reduces the relapse rate in steroid-dependent nephrotic syndrome (SDNS), studies on comparative efficacy are lacking. We retrospectively reviewed the records of patients with difficult-to-treat SDNS who had previously received levamisole, cyclophosphamide and/or mycophenolate mofetil, then treated with either rituximab or tacrolimus and followed for 12 months. Between January 2009 and April 2010, ten patients received two to three doses of IV rituximab (375 mg/m(2)/week) and 13 received tacrolimus (0.1-0.2 mg/kg/day) for 12 months; none had previously received either agent. Patients received tapering doses of alternate-day prednisolone; other immunosuppressive agents were discontinued. The mean age of the patients at treatment initiation with rituximab and tacrolimus was 12.2 ± 2.3 and 12.3 ± 3.0 years, respectively. The respective pre-treatment relapse rates (3.1 ± 1.1 and 3.5 ± 1.6 relapses per year) and cumulative prednisolone dose (137.2 ± 69.4 and 140.5 ± 59.0 mg/kg/year) were similar. Therapy resulted in a decline in relapse rate in both groups (P < 0.001). The number of relapses in the rituximab and tacrolimus groups was similar at 6 months (0.3 ± 0.5 vs. 0.3 ± 0.6 episodes, respectively), 12 months (0.8 ± 1.0 vs. 0.9 ± 1.1 episodes) and last follow-up (1.2 ± 1.0 vs. 1.5 ± 1.3 episodes). There were no differences in relapse-free survival at 6, 12 and 18 months. Therapy resulted in a significant decline in the cumulative prednisolone dose (67.2% in the rituximab group and 43.6% in the tacrolimus group) and a reduced body mass index. These findings suggest that in our patients with difficult-to-treat SDNS, treatment with two to three doses of rituximab was as effective as 12 months of therapy with tacrolimus in terms of steroid sparing and reduction in the relapse rate.