Increasing rates of invasive candidiasis caused by non-albicans Candida species have been reported worldwide. Particular concerns have been raised for C. parapsilosis because of reduced in vitro susceptibility to echinocandins. We identified 212 patients with invasive candidiasis due to non-albicans Candida species (>or=5 cases per species) in 5 clinical trials of caspofungin monotherapy from the pharmaceutical sponsor's (Merck and Co., Inc.) database: 71 cases were caused by C. parapsilosis, 65 by C. tropicalis, 54 by C. glabrata, 10 by C. krusei, 9 by C. guilliermondii, and 5 by C. lusitaniae. One hundred sixty-seven cases caused by C. albicans were also identified. Efficacy was assessed at the end of caspofungin therapy. Success (favorable overall response) required favorable clinical and microbiological responses. The mean APACHE II scores were 16.5 in the non-albicans group and 15.7 in the C. albicans group. Neutropenia at study entry was more common in the non-albicans group (12%) than in the C. albicans group (5%). The median duration of caspofungin therapy was 14 days in both groups. The success rates were 77% in both groups and at least 70% for each non-albicans species: 74% for C. parapsilosis, 71% for C. tropicalis, 85% for C. glabrata, 70% for C. krusei, 89% for C. guilliermondii, and 100% for C. lusitaniae. The times to negative blood culture were similar for the various species. The overall mortality rates were 26% in the non-albicans group and 29% in the C. albicans group. Drug-related serious adverse events and discontinuations due to caspofungin toxicity were uncommon. Although the sample sizes were limited, caspofungin demonstrated favorable efficacy and safety profiles in the treatment of invasive candidiasis caused by the following non-albicans Candida species: C. parapsilosis, C. tropicalis, C. glabrata, C. krusei, C. guilliermondii, and C. lusitaniae.