Background: The U.S. Food and Drug Administration's (FDA) Bacteriological Analytical Manual (BAM) uses Bacillus cereus rapid agar (BACARA) and Mannitol-yolk-polymyxin (MYP) agar for the enumeration of the members of B. cereus group. Objective: The automated TEMPO Most Probable Number system was compared with the FDA BAM method for the detection of B. cereus group members in cosmetic products. Methods: We inoculated a range of cosmetic products with pure B. cereus spore suspensions (density = 0.5 McFarland) at high (6 log CFU/mL), medium (5 log CFU/mL), and low (4 log CFU/mL) levels. Test portions were aged for 72 h. Five replicates per sample were analyzed; uninoculated test portions served as controls. We also evaluated whether TEMPO BC erroneously detected non-B. cereus or other adulterant organisms. Results: No significant differences (P > 0.05) were found among the TEMPO BC and the BAM spiral plating methods. Correlations between TEMPO BC - BACARA and TEMPO BC - MYP were 0.895 and 0.893 for powder type products, 0.834 and 0.846 for cream and oil-based products, and 0.929 and 0.923 for liquid products, respectively. Non-B. cereus strains were not detected by TEMPO BC. Conclusions: The TEMPO BC method can be used for the detection of B. cereus in cosmetic products without preservatives, or those preserved with either phenoxyethanol or other organic substances.