In this study, we evaluated the ability of a 4-h enzyme assay kit system, the RapID ANA method (Innovative Diagnostic Systems, Inc., Atlanta, Ga.) to accurately and reproducibly identify a spectrum of clinically significant anaerobic bacteria in two separate institutions. Additional tests were performed as required. Of a total of 188 organisms tested at Hershey Medical Center (HMC), 86.2% were correctly identified to species level without additional tests, 5.9% required extra tests for correct identification, and 8.0% were misidentified. Of 53 strains tested at Johns Hopkins Hospital (JHH), 52.8% were correctly identified without extra tests, 28.3% required extra tests for correct identification, and 18.9% were misidentified. Of 21 organisms tested at both institutions, those tested at JHH required additional tests for correct identification in 38.1% of cases, compared with 9.5% at HMC. Misidentification rates were identical (9.5%) in both centers. Of strains tested at HMC only, 86.8% were correctly identified without extra tests, 5.4% were identified with additional tests, and 7.8% were misidentified: corresponding data for JHH were 53.1, 21.9, and 25.0%, respectively. Of 53 strains tested in triplicate at JHH, 56.7% yielded the same result on each occasion, 37.7% were identical in two of three tests, and 5.7% gave different results on each of three occasions. Discrepancies between identification rates at HMC and JHH may be explained by differences in species tested (more commonly encountered species were tested at HMC) and interpretation of reactions by the two different readers. The RapID ANA method has the potential for rapid identification of clinically isolated anaerobes; however, accuracy and reproducibility may vary as a function of the specific laboratory setting.