The grouping of beta-hemolytic streptococcal isolates by a new direct-plate procedure employing Phadebact Streptococcus Test reagents was compared with the results obtained with the 4- and 24-h Phadebact grouping procedure and with the Lancefield grouping obtained with a capillary precipitin test. The new procedure employed a modification of the Phadebact procedure that permitted the grouping of streptococci on glass slides with a minimum of five primary isolated colonies. When only five to eight colonies were available for direct testing with each Phadebact reagent, coagglutination was better manifested when the colonies were disaggregated on a glass slide in a loopful of Tween 80 solution. Further enhancement of the coagglutination reaction was effected when the respective Phadebact reagents were employed in relatively small volumes. The direct-plate procedure permitted the correct identification of 127 out of 129 betahemolytic isolates. The 4-h method correctly identified 192 of the 200 streptococci tested. All of the 200 isolates tested by the 24-h procedure and the Lancefield grouping were correctly identified. The direct-plate Phadebact procedure affords the clinical microbiologist a rapid and reliable means of identifying groups A, B, C, and G beta-hemolytic streptococci. When sufficient numbers of primary colonies are not available for the direct procedure, the 4- or 24-h procedures may be employed.