An evaluation of the Vitek Anaerobe Identification (ANI) card was performed with 341 bacterial isolates, including 313 clinical isolates and 28 stock strains of anaerobic microorganisms. Identifications obtained with the ANI card were compared with those determined by conventional methods. The card identified 73.2% of 149 anaerobic gram-negative bacilli, 63.6% of 44 Clostridium spp., 65.8% of 38 anaerobic nonsporeforming gram-positive bacilli, and 69.1% of 110 anaerobic cocci, with no further testing required. When genus-level identifications were included, 83.9% of the anaerobic gram-negative bacilli, 70.5% of Clostridium spp., 73.7% of the anaerobic nonsporeforming gram-positive bacilli, and 73.6% of the anaerobic cocci were identified. Nineteen isolates (5.6%) produced identifications of good confidence but marginal separation or questionable biotype, in which the correct identification was listed with one or two other possible choices and extra tests were required and suggested. A total of 28 (8.2%) were not identified and 29 isolates (8.5%) were misidentified by the ANI card. Among the commonly isolated clinically significant anaerobes, the ANI card identified 100% of 55 Bacteroides fragilis and 100% of 8 Clostridium perfringens. Use of supplemental tests and expansion of the data base to include additional strains of organisms that are difficult to separate even with conventional methods may improve the accuracy of the ANI card as a method for identification of anaerobic bacteria in the clinical laboratory.