We conducted a retrospective study to evaluate the duration of optimal steroid therapy in idiopathic focal segmental glomerulosclerosis (FSGS). We evaluated 93 adult patients (n=65 males) with biopsy proven FSGS. Mean proteinuria was 5.4 +/- 2.8 gm/dL. Twelve patients were lost at follow-up. Of the remaining 81 patients, nephrotic range proteinuria was present in 48 (59%), and 21 (26%) presented with renal insufficiency. Of these patients, three (3.9%) experienced spontaneous remission. Seven patients were managed symptomatically with ACE inhibitors and never received steroids. Of the 71 patients, 32 received >16 weeks of steroid therapy, while 39 received <16 weeks of steroid therapy. Twenty-four patients (75%) who received >16 weeks of steroid therapy had a complete or partial remission, while only 18 (46%) of those with <16 weeks of steroid therapy had a steroid response (p=0.001). Patients with more than 25% interstitial fibrosis at biopsy also showed significantly lower remission rates (p=0.02). Hypertension, hematuria and degree of proteinuria did not significantly affect the response to steroid therapy. Univariate logistic regression analysis showed that the factors predictive of remission were: (1) steroid therapy duration (p=0.001); (2) serum creatinine (Cr) at onset (p=0.001) and; (3) presence of interstitial fibrosis (>25%) at initial biopsy (p=0.02). Multivariate logistic regression analysis showed that the only factor predictive of remission was steroid therapy duration >16 weeks (p=0.001). Therefore, we concluded that patients with idiopathic FSGS required treatment for at least 16 weeks, before labeling them as steroid non-responsive. Patients with interstitial fibrosis have a significantly poor response to therapy.