AIMS: To compare the performance of seven commercial yeast identification methods with that of a reference method, and to compare the costs of the commercial kits. METHODS: Clinical yeast isolates (n = 52), comprising 19 species, were identified using Vitek, Api ID 32C, Api 20C AUX, Yeast Star, Auxacolor, RapID Yeast Plus system, and Api Candida and compared with a reference method which employed conventional tests. RESULTS: The percentage of correctly identified isolates varied between 59.6% and 80.8%. Overall, the highest performance was obtained with Api Candida (78.8%) and Auxacolor (80.8%). Among germ tube negative yeast isolates, Auxacolor and Api Candida both identified 93.1% of isolates correctly. All systems failed to identify C norvegensis, C catenulata, C haemulonii, and C dubliniensis. In comparison with Auxacolor, the Api Candida is less expensive and requires less bench time. CONCLUSIONS: Auxacolor and Api Candida appeared to be the most useful systems for identification of germ tube negative yeast isolates in clinical microbiology laboratories, although one should be aware that several germ tube negative Candida species cannot be identified by these systems.