Recurrence of steroid-resistant nephrotic syndrome in kidney transplants is associated with increased acute renal failure and acute rejection. We performed 73 kidney transplants in 51 patients with steroid-resistant nephrotic syndrome (SRNS) with focal segmental glomerular sclerosis (FSG) ages 18.4 ± 12.8 (X ± sd) years. Recurrence of SRNS, defined by rapid onset of proteinuria, hypoalbuminemia and/or >95% epithelial cell foot process effacement with or without the presence of FSG, occurred in 26 grafts in 16 patients. Acute renal failure (ARF) occurred in 16 of 26 (61.5%) grafts with recurrence versus 7 of 47 (14.9%) grafts without recurrence (P < 0.0001). ARF occurred in 4 of 9 (44.4%) living-related donor (LRD) recipients with recurrence and 3 of 21 (12.5%) LRD recipients without recurrence (NS). ARF in cadaver donor (CAD) recipients with recurrence was 12 of 17 (70.5%) versus 4 of 23 (17.4%) without recurrence (P < 0.0001). ARF was also higher in LRD or CAD with recurrence than in a control group of non-SRNS patients matched for age, sex and time of transplantation. Graft survival at one year was lower in patients with recurrence and ARF [4 of 16 (25%)] compared to patients with recurrence and no ARF [9 of 11 (82%), P < 0.01]. There was no difference in graft survival in patients without recurrence who did or did not have ARF. One or more acute rejection episodes occurred in all 16 patients with ARF and recurrence, in all 7 patients with ARF without recurrence, and in 7 of 10 patients with recurrence without ARF compared with only 11 of 40 (28%) of patients with neither recurrence nor ARF (P < 0.0001, <0.001 and <0.04, respectively). A loglinear model showed associations between acute rejection and ARF (P < 0.01), recurrence and ARF (P = 0.05), and recurrence and acute rejection (P < 0.025). Twelve of 16 (75%) of grafts with both recurrence and ARF versus 4 of 40 (10%) of grafts with neither recurrence nor ARF were lost in association with acute rejection in the first two post-transplant years (P < 0.0001). Only one graft was lost from recurrence of SRNS, per se. Thus, recurrence of SRNS is associated with an increased incidence of ARF, and both ARF and recurrence are associated with greater risks of acute rejection. The increased graft loss in recurrence is primarily due to acute rejection. As has been speculated for ARF, recurrence may result in graft changes which stimulate acute rejection.