A biphasic blood culture bottle (BiPB: GIBCO Laboratories, North Andover, Mass.) with an architectural design that physically separates the agar slant from the broth was compared with a conventional vented monophasic bottle (MPB-A) for use in the routine culture of blood. Both bottles contained tryptic soy broth. Tryptic soy agar was used for the BiPB slant. A third unvented bottle (MPB-N) with Columbia broth was included as part of the blood culture set. Of 3,537 sets collected, 444 were positive; 57 of these 444 sets were positive by virtue of an exclusively positive anaerobic bottle. Both BiPB and MPB-A were positive in 235 of the remaining 387 positive sets. A total of 521 isolates was recovered during the study. Of these isolates, 252 were recovered in both the BiPB and the MPB-A from the same set; 105 isolates grew in the BIPB but not in MPB-A, 95 isolates grew only in the MPB-A but not in BiPB, and 69 grew exclusively in the MPB-N. The BiPB allowed more rapid recovery of Candida spp., J-K diphtheroids, Pseudomonas spp. Making BiPB subcultures was easy enough to permit both early and daily subculture, which provided isolated colonies sooner than could be done by using the MPB-A. Isolated colonies and, therefore, identification and susceptibility results were available at least 1 day earlier for the BiPB isolates in approximately 50% of instances when both the BiPB and the MPB-A were positive. Staphylococcus epidermidis and streptococci were recovered more frequently in the BiPB, while gram-positive anaerobes were detected at a significantly (P less than 0.025) more frequent rate in the MPB-A than in the BiPB. Either bottle, however, should be used in conjunction with an anaerobic bottle for optimal recovery of anaerobic bacteria.