Vacuum filter socks were evaluated for recovery efficiency of powdered Bacillus atrophaeus spores from two non-porous surfaces, stainless steel and painted wallboard and two porous surfaces, carpet and bare concrete. Two surface coupons were positioned side-by-side and seeded with aerosolized Bacillus atrophaeus spores. One of the surfaces, a stainless steel reference coupon, was sized to fit into a sample vial for direct spore removal, while the other surface, a sample surface coupon, was sized for a vacuum collection application. Deposited spore material was directly removed from the reference coupon surface and cultured for enumeration of colony forming units (CFU), while deposited spore material was collected from the sample coupon using the vacuum filter sock method, extracted by sonication and cultured for enumeration. Recovery efficiency, which is a measure of overall transfer effectiveness from the surface to culture, was calculated as the number of CFU enumerated from the filter sock sample per unit area relative to the number of CFU enumerated from the co-located reference coupon per unit area. The observed mean filter sock recovery efficiency from stainless steel was 0.29 (SD = 0.14, n = 36), from painted wallboard was 0.25 (SD = 0.15, n = 36), from carpet was 0.28 (SD = 0.13, n = 40) and from bare concrete was 0.19 (SD = 0.14, n = 44). Vacuum filter sock recovery quantitative limits of detection were estimated at 105 CFU m(-2) from stainless steel and carpet, 120 CFU m(-2) from painted wallboard and 160 CFU m(-2) from bare concrete. The method recovery efficiency and limits of detection established in this work provide useful guidance for the planning of incident response environmental sampling for biological agents such as Bacillus anthracis.