BACKGROUND Automated culture methods have been used by several investigators to detect bacterial contamination of cellular blood components. We investigated several factors affecting detection by automated culture of bacteria in platelet concentrates (PCs). These factors included the initial contamination level in PCs, the PC sample volume, the PC sample time, and the white cell level in relation to bacteria levels in the PCs. METHODS Staphylococcus epidermidis or Escherichia coli was inoculated into freshly prepared PCs or white cell-reduced PCs to yield colony-forming unit (CFU) levels of 10, 1, or 0.1 per mL. At the time of inoculation (t=0) and at t=6, t=24, and t=48 hours, 0.5, 1.0, and 2.0 mL samples of the contaminated PCs were transferred into culture bottles. The presence of bacteria in the culture bottles was subsequently monitored by an automated blood culturing instrument. Bacteria levels in the PC at the time of first automated culture detection were determined by quantitative plating. RESULTS E. coli was detected in 92 percent of experiments when 1.0- or 2.0-mL samples were taken at t=6 hours. At t=24 hours, 100-percent detection was observed with all tested inoculation volumes; however, by that time, >10(7) CFU per mL of bacteria were present in every PC. For S. epidermidis, 89 percent and 83 percent of contaminated PCs were detected with a t=24 hour sampling time and 2.0- or 1.0-mL sampling volume. Seven of 36 PCs with a 2.0-mL sampling volume and 10 of 36 PCs with a 1.0-mL sampling volume contained >10(6) CFU per mL of S. epidermidis at the time of first detection. CONCLUSIONS Data from this preliminary evaluation suggest that sampling times of 24 hours or more would be necessary to provide confidence in detection of E. coli or S. epidermidis in PCs using this culture method.