The Andersen Microbial Sampler (AMS) has been used under varying conditions, bur it has not been fully characterized with respect to the effect of sampling time on total sampler efficiency. The objective of this study was to characterize the AMS with respect to sampling time on total efficiency, measured as colony-forming units per cubic meter (CFU m(-3)) when sampling for total culturable bacteria with Tryptic-soy agar (TSA) at an outdoor compost center. The sample times ranged from 1 to 6 min. Statistical analysis using multivariate analysis of variance (MANOVA) coupled with mixed model analysis of variance showed a parabolic trend between CFU m(-3) and sampling run rime. At both sites, the highest average CFU m(-3) occurred at a sampling run time of 2 min. Duplicate samples were taken, and a test of internal validity was performed to show that the sampling methodology used was not a source of error, The Pearson＇s correlation coefficient for the two samplers was 0.86 at the first site and 0.90 at the second. The: effect of sampling time on the particles impacting the second stage was also examined and found not to be significant. Even though field testing presents uncontrollable variables, optimal sampling run times can still be determined using statistical analyses. Thus, optimal sampling times based on a complex microflora and not a single genus will minimize unsuccessful sampling episodes (e.g. plate overloading).