The performance of the RapID Yeast Plus System (Innovative Diagnostic Systems, Norcross, Ga.), a 4-h micropanel using single-substrate enzymatic test reactions, was compared with that of the API 20C AUX Clinical Yeast System (bioMerieux Vitek, Hazelwood, Mo.), a 48- to 72-h carbohydrate assimilation panel. Two hundred twenty-five yeasts, yeast-like fungi, and algae, comprising 28 species and including 30 isolates of Cryptococcus neoformans, an important pathogen not tested in appreciable numbers in other comparisons, were tested by both methods. On initial testing, 196 (87.1%) and 215 (95.6%) isolates were correctly identified by the RapID and API systems, respectively. Upon repeat testing, the number of correctly identified isolates increased to 220 (97.8%) for the RapID system and 223 (99.1%) for the API system. Reducing the turbidity of the test inoculum to that of a no. 3 McFarland turbidity standard, which is below that recommended by the manufacturer, resulted in the correct identification of most of the isolates initially misidentified by the RapID system, including 10 of 30 C. neoformans isolates. Concordance between the RapID and API results after repeat testing was 97.3%.