Background: Intravenous cyclosporine is active in 60% to 80% of patients with ulcerative colitis (UC) who failed to respond to intravenous corticosteroids. Several studies have suggested that cyclosporine in microemulsion form (Neoral) has some efficacy in this setting, but the optimal dose, blood level, time to response, and remission need to be better defined. The aim of this study was to evaluate the response to Neoral and its toxicity in active corticosteroid-refractory UC. Methods: Between March 2002 and August 2005, 20 courses of Neoral [initial dose, 2.3 mg/kg (range, 1.8 to 2.8 mg/kg) every 12 hours] were prescribed in 19 consecutive patients for a UC attack that did not respond to intravenous methylprednisolone. All patients received prophylaxis against Pneumocystis carinii. Results: Response was obtained in 17 of 20 attacks (85%) after 3.5 days (range, 1 to 7). Remission was obtained in 15 of 20 attacks (75%) after 13 days (range, 2 to 30 days). Four responders relapsed and underwent colectomy 21 to 900 days after the start of Neoral. Overall, 14 of 19 patients (74%) were colectomy free after a median follow-up of 8 months (range, 1 to 41 months). Cyclosporine blood levels were measured at fasting (C0) and 2 hours after Neoral administration (C2) in a subgroup of 10 responders. The results were 103 ng/mL (range, 32 to 240 ng/mL) for C0 and 761 ng/mL (183 to 1390 ng/mL) for C2. One severe bedridden patient with neonatal encephalopathy died. Main side effects observed were mild transient renal impairment (n = 2), hypertension (n = 1), cytomegalovirus infection (n = 2), and esophageal candidiasis (n = 1). Conclusions: In active corticosteroid-refractory UC, Neoral seems to have the same efficacy and toxicity as the intravenous form. Trough target cyclosporine blood levels should not exceed 100 ng/mL for C0 and 700 ng/mL for C2.