PURPOSE: We examined, retrospectively, the efficacy of voriconazole in Fusarium eye infections. METHODS: Voriconazole-treated patients with proven or probable keratitis or endophthalmitis from the voriconazole database (9 patients) and six French ophthalmology departments (15 patients) were included. Sociodemographic features, predisposing factors, history of corneal trauma, associated ocular conditions, other diseases and prior therapies were analysed. Investigator-determined success was defined as infection resolution with medical treatment. Failure was no response or persistent infection and required surgery. RESULTS: Most patients were Caucasian (83 %) and male (71 %). The infection was keratitis (63 %) or endophthalmitis (37 %) and proven in 23 (96 %). Prior therapy included topical and/or systemic amphotericin (46 %), fluconazole (17 %) or others (33 %), often in combination. Causative fungi were Fusarium solani (14, 58 %), Fusarium moniliforme (1), Fusarium oxysporum (1) and Fusarium spp. (8). Voriconazole was administered systemically, topically and/or by intraocular injection, and 16 patients (67 %) received salvage and eight primary therapy. The overall response was 67 % (73 % keratitis and 56 % endophthalmitis) but seven patients required adjunctive surgery. However, response was 63 % for eight primary therapy patients and 69 % for 16 salvage therapy patients. Response by species was Fusarium solani 64 % (9/14) and all others 80 % (8/10). In 13 patients (77 %), voriconazole was used in combination (response 69 vs. 64 % alone) with topical [amphotericin B 10/24 (42 %), caspofungin 5 (21 %), natamycin 1 (4 %)] and systemic agents [caspofungin 3 (13 %), amphotericin 2 (8 %)]. CONCLUSIONS: Topical and systemic voriconazole appears to be effective alone or in combination with other agents for treating severe Fusarium keratitis or endophthalmitis.