We placed a Labsystems Bioscreen C instrument in an anaerobic chamber to assess inoculum effects on broth microdilution testing in the first hours of bacterial growth rather than on MIC values resulting after 48h. The Bioscreen C can automatically dispense broth, antibiotics, and inoculum, then incubate and spectrophotometrically monitor growth. Organisms (Bacteroides fragilis ATCC 25285 and Bacteroides thetaiotaomicron ATCC 29741) were tested in brucella broth (BB) with meropenem (Mer), ampicillin/sulbactam (A/S,2:1), and metronidazole (Met). Cells from overnight brucella blood agar cultures were suspended in BB using a Vitek colorimeter and then diluted 1:4 in PRAS dilution blanks. Changes in inoculum size did not cause large differences in growth curves for control wells. For B. fragilis inoculated at 6.2X105, initiation of log (5h) phase occurred 2h later than when the inoculum was 3.7X106. For B. thetaiotaomicron inoculated at 9X105, initiation of log (5h) occurred 1h later than when inoculated at 3.4X106. In wells containing antimicrobial agents, inoculum effects were much greater. In tests of B. fragilis with Mer, midpoints of log phase occurred at approximately 7h and 27-29h at 0.03 æg/ml and 0.06 æg/ml of Mer, respectively, when inoculated at 3.7X106. However, when the inoculum size of B. fragilis was 1.3X106, midpoints of log phase occurred at approximately 20-23h and >48h for 0.03 æg/ml and 0.06 æg/ml of Mer, respectively. Midpoints of log phase occurred at 8h and 18h at concentrations of 0.12 and 0.25 æg/ml of Met, respectively, for a B. fragilis inoculum of 3.8X106 compared to 34h and >48h for an inoculum of 6.2X105. In tests of B. thetaiotaomicron with 0.5 æg/ml A/S, midpoints of log phase ranged from 30h to 48h for inocula of 2.8X106 and 9X105, respectively. Although the onset of log phase growth was substantially affected by small changes in inoculum size, microdilution MIC values were maintained within one twofold dilution after 48h. Thus, differences in inoculum effects were masked by an inappropriately long incubation period for these fast growing anaerobes. While this incubation period may provide a consensus among laboratories by minimizing differences and experimental errors, possible clinically relevant differences in antimicrobial activity were no longer evident.