A total of 200 beta-hemolytic streptococci, isolated from clinical specimens submitted to our laboratory, were identified as group A versus non-A using the fluorescent antibody technique (FA), bacitracin susceptibility (BBL, Difco, and Raven disks), SeroSTAT, Streptex, Phadebact, and the API 20S system. Of the 122 group A isolates, all methods except SeroSTAT and Phadebact yielded 92-99% agreement when compared with the Lancefield precipitin test. Phadebact yielded an 84% agreement and SeroSTAT changed from 83 to 98% after trypsinization. Numerous false positives were obtained and only FA (91%) and API 20S (96%) yielded better than 90% agreement on non-A identification when compared with the Lancefield test. The most false positives were obtained (45%) using the SeroSTAT reagents. Considering accuracy, our data suggests the FA technique to be the method of choice for identifying group A streptococci.