Meritec-Strep (Meridian Diagnostics, Inc., Cincinnati, Ohio) a coagglutination method, was compared with Streptex (Wellcome Diagnostics, Research Triangle Park, N.C.), a latex agglutination method, for the identification of beta-hemolytic streptococcal groups A, B, C, F, and G by the direct colony method. A total of 124 beta-hemolytic streptococcal isolates were tested, which included 77 from group A, 15 from group B, 10 from group C, 1 from group F, and 21 from group G. All were tested from subculture, and 74 (60%) were also tested from primary isolation plates. For Meritec-Strep, usually one colony was directly applied to the reaction card for testing each grouping reagent, while for Streptex, five colonies were tested after a 1-h extraction process. Complete agreement was obtained for all isolates tested from subculture with the kits. From primary isolation plates, Meritec-Strep correctly identified 97.3% of the isolates compared with 94.6% correctly identified for Streptex. Meritec-Strep produced a false-negative for one group A isolate and positive reactions for group A and F reagents with another group A isolate. A diphtheroid contaminant caused the positive group F reaction. Streptex produced false-negative results for one group A and three group C isolates. Most positive reactions were strong and rapid (less than 30 s) for both kits. The negative test control provided in the individual group A and B kits was nonreactive for all isolates. Meritec-Strep accurately identified isolated colonies of beta-hemolytic streptococci on primary isolation and subculture plates. It provided faster results than Streptex by eliminating the time and manipulation of antigen extraction and needed fewer colonies when individual group A or B reagents were used.