The authors performed a retrospective clinicopathologic study in 81 patients with primary focal segmental glomerular sclerosis (FSGS) to determine whether they could identify clinical or histologic features at presentation that could be predictive of outcome and response to therapy. Males constituted 58% of patients, and 53% were black. At biopsy the patients were 40 ± 17 years old; 74% were nephrotic, and renal insufficiency was present in 62%. The average time from presentation to biopsy was 16 months, and the average total follow-up was 62 months. Nephrotic patients had a significantly poorer prognosis as compared with nonnephrotic patients (5- and 10-year survivals of 76% and 57% v 92% and 92%; P < 0.05). A multivariate analysis was done on histologic and clinical features at biopsy, assessing for risk factors leading to end-stage renal disease, showing only the serum creatinine and the degree of interstitial fibrosis to have a significant correlation. Thirty nephrotic patients received prednisone, with a treatment time of 5.5 ± 4 months and a total dose of 5.9 ± 2.9 g per course of treatment. Fifteen of these patients (50%) achieved a remission by 3.7 ± 2 months (10 complete remission and 5 partial remissions), with all patients responding within 9 months. Only two patients had spontaneous remissions (both partial). The 5- and 10-year survival for patients in remission were both 100% as compared with 66% and 41% (P < 0.01), respectively, for nephrotic patients not in remission. No clinical feature at presentation or biopsy was predictive of response to therapy when a multivariate analysis was performed. Nephrotic patients responding to therapy have a significantly better prognosis than nonresponsive patients. Response to therapy requires at least 3 to 4 months of treatment with prednisone. Unfortunately, no clinical feature at presentation or biopsy predicts which patients are more likely to respond to treatment. This experience suggests that nephrotic patients with primary FSGS may benefit from a more prolonged course of therapy with prednisone. Furthermore, these findings underscore the need for a controlled trial in nephrotic patients with FSGS to confirm the response to and establish guidelines for therapy.