BACKGROUND: It is well established that automated blood culture systems require no more than five days of incubation for the detection of the majority of pathogens. It is not clear, however, whether continuous monitoring of blood culture systems also routinely require five days of incubation. This study was conducted to determine the clinical impact of incubating blood cultures for 4 days rather than for 5 days using the BACTEC 9240 blood culture system. METHODS: During the 6-month period from July to November 1998, 22,167 blood cultures were performed. Positive culture sets and the isolates were sorted by times to detection of isolates. Chart reviews were done for isolates detected on day 3 or later to determine whether therapy was changed due to this blood culture result. RESULTS: Of 2,426 isolates (2,319 positive cultures), 2,344 (96.6%) were recovered within 3 days and 52 (2.1%) were recovered on day 4, and 30 (1.2%) on day 5. Chart reviews showed that 21 of the 52 isolates detected on day 4 were considered clinically significant and 10 of those affected the treatment of the patients. On day 5, 5 of the 30 isolates were considered clinically significant and 3 of those affected the treatment. CONCLUSIONS: Four-days rather than a 5-day incubation period reduced culture sensitivity by 1.2% but most of those were clinically irrelevant. These data suggest that the 4-day protocol for the BACTEC 9240 system is adequate for detection of positive blood cultures.