We tested for soil substrate effects on the movement and infectivity of naturally co-occurring entomopathogenic nematodes Steinernema feltiae and Heterorhabditis marelatus, alone and in combination. We manipulated the presence and bulk density of soil and added Galleria mellonella baits within capped and perforated 15 mL centrifuge tubes. Sampling tubes were then deployed in situ into field and laboratory settings as experimental traps for infective juveniles. In comparisons with standard soil collections from Lupinus arboreus rhizospheres, sampling tubes were equally sensitive to the presence of H. marelatus and more sensitive to S. feltiae. In laboratory microcosms, both EPN species infected Galleria at high frequencies in tubes lacking soil and in the absence of heterospecifics. Infection frequency of S. feltiae was unaffected by the presence of H. marelatus, but it declined with higher soil bulk density inside tubes. In contrast, detectable infection frequency by H. marelatus was reduced only marginally by the presence of soil but severely by the presence of S. feltiae. Thus, the presence of soil in tubes reversed the identity of dominant species infecting Galleria in tubes, an effect magnified when soils were compacted. Moreover, S. feltiae rarely moved into tubes lacking Galleria baits, whereas H. marelatus colonized unbaited tubes 4- to 5-fold more frequently than S. feltiae. In situ, sampling tubes acted as filters to reduce interference and contamination by fungal pathogens common in field soils. The method allows precision sampling with minimal soil disturbance while protecting bait insects from scavengers. Manipulation of tube design may allow selective sampling of EPN species, depending on the abiotic characteristics of soils, and the biology, behavior, and interspecific interactions of coexisting species.