We sought to evaluate the effects of Bacillus species, low surface concentrations, and surface material on recovery efficiency (RE), false-negative rate (FNR) and limit of detection for recovering Bacillus spores using a validated macrofoam-swab sampling procedure. The performance of a macrofoam-swab sampling method was evaluated using Bacillus anthracis Sterne (BAS) and Bacillus atrophaeus Nakamura (BG) spores applied at nine low target surface concentrations (2 to 500CFU per plate or coupon) to positive-control plates and test coupons (25·8064cm(2) ) of four surface materials (glass, stainless steel, vinyl tile and plastic). The Bacillus species and surface material had statistically significant effects on RE, but surface concentration did not. Mean REs were the lowest for vinyl tile (50·8% with BAS and 40·2% with BG) and the highest for glass (92·8% with BAS and 71·4% with BG). FNR values (which ranged from 0 to 0·833 for BAS and from 0 to 0·806 for BG) increased as surface concentration decreased in the range tested. Surface material also had a statistically significant effect on FNR, with FNR the lowest for glass and highest for vinyl tile. Finally, FNR tended to be higher for BG than for BAS at lower surface concentrations, especially for glass. Concentration and surface material had significant effects on FNR, with Bacillus species having a small effect. Species and surface material had significant effects on RE, with surface concentration having a nonsignificant effect. The results provide valuable information on the performance of the macrofoam-swab method for low surface concentrations of Bacillus spores, which can be adapted to assess the likelihood that there is no contamination when all macrofoam-swab samples fail to detect B.anthracis.