A liquid culture enrichment-polymerase chain reaction (E-PCR) assay was investigated as a potential tool to overcome inhibition by chemical component, debris, and background biological impurities in soil that were affecting detection assay performance for soil samples containing Bacillus atrophaeus subsp. globigii (a surrogate for B. anthracis). To evaluate this assay, 9 g of matched sets of three different soil types (loamy sand [sand], sandy loam [loam] and clay) was spiked with 0, ~4.5, 45, 225, 675 and 1350 endospores. One matched set was evaluated using a previously published endospore concentration and colony-forming unit spreadplate (CFU-S) assay and the other matched set was evaluated using an E-PCR assay to investigate differences in limits of detection between the two assays. Data illustrated that detection using the CFU-S assay at the 45-endospore spike level started to become sporadic whereas the E-PCR assay produced repeatable detection at the ~4.5-endospore spike concentration. The E-PCR produced an ~2-log increase in sensitivity and required slightly less time to complete than the CFU-S assay. This study also investigated differences in recovery among pure and blended sand and clay soils and found potential activation of B. anthracis in predominately clay-based soils.