The method detection limit (MDL, 99% chance of detecting a positive result in a single replicate), as per the United States Code of Federal Regulations, was determined for a protocol using an ultrafiltration based automated waterborne pathogen concentration device. Bacillus anthracis Sterne strain spores were seeded at low levels into 100 L reagent water samples. Suspect colonies were confirmed through morphological, chemical, and genetic tests. Samples of 100 L (n=14) of reagent water were seeded with five B. anthracis CFUs each. To confirm the estimated detection limit, a second set (n=19) of 100 L reagent water samples were seeded at a higher level (7 CFUs). The second estimate of the MDL could not be pooled with the first, due to significant difference in variance. A third trial (n=7) seeded with 10 CFUs produced an estimate of the MDL that could be pooled with the higher previous estimate. Another trial consisting of eight 100 L samples of tap water were seeded with approximately 7 CFUs. Recovery in these samples was not significantly different from the pooled MDL. Theoretically a concentration of 4.6 spores/100 L would be required for detection 95% of the time, based on a Poisson distribution. The calculated pooled MDL, based on experimental data was approximately 6 B. anthracis CFU/100 L (95% confidence interval 4.8 to 8.4). Detection at this level was achieved in municipal water samples.