We compared the Difco ESP 384 blood culture system with the pediatric Septi-Chek system for the detection of bloodstream infections in pediatric patients. A total of 10,762 blood culture sets included an ESP aerobic bottle and a Septi-Chek bottle, From these cultures, a total of 278 organisms classified as probable pathogens were isolated, including 237 from ESP bottles and 221 from Septi-Chek bottles. This difference was not statistically significant. More organisms classified as possible contaminants were also isolated from ESP bottles (for ESP, 480 bottles; for Septi-Chek, 418 bottles; P < 0.01). The time to detection was shorter for probable pathogens isolated from ESP bottles (median times for organisms isolated from both Systems: ESP, 14.0 h; Septi-Chek, 34.5 h; P < 0.001. The proportions of all probable pathogens detected by 24 and 48 h after inoculation were 78 and 96%, respectively, for ESP compared with 31 and 74%, respectively, for Septi-Chek The time to final identification was also shorter for organisms grown in ESP bottles (median times for organisms isolated from both systems: ESP, 48.8 h; Septi-Chek, 58.5 h; P less than or equal to 0.001). A subset of 4,442 cultures also included an ESP anaerobic bottle in addition to an ESP aerobic bottle and a Septi-Chek bottle. There were no significant differences in the recovery of probable pathogens by any of the possible two bottle combinations, but five anaerobic pathogens were recovered;only in the anaerobic bottle. We conclude that the ESP 384 is an excellent system for culturing pediatric blood samples and that it provides for the very rap;detection of bloodstream pathogens.