Cobble baskets are frequently used as a tool to measure differences in benthic macroinvertebrate communities between waterbodies; however, underlying differences in substrate type may influence the resultant colonization of baskets, misrepresenting communities. This study tests the hypothesis that cobble basket placement influences the resulting benthic macroinvertebrate community. Cobble basket arrays (n = 4) were deployed in Dog Lake, Saskatchewan, in 2011 (97 d) and 2012 (95 d) on cobble habitats and soft or sandy substrates ~100 m apart. Baskets placed on cobble substrate had significantly higher Shannon–Weaver diversity relative to those placed on soft substrate in both years, and higher % EPT (Ephemeroptera Plecoptera Trichoptera) in 2011, but total density was not significantly different. Nonmetric multidimensional scaling revealed that the community was different between both treatments, characterized by higher densities of Gammarus lacustris Sars in baskets placed on soft sediment in both years, higher densities of Aeshna sp. and Mystacides sp. on cobble substrate in 2011, and higher densities of Helobdella stagnalis (L.) and Glossophinia complanata (L.) on cobble substrate in 2012. The results were consistent with the hypothesis that baskets placed on cobble substrate versus soft substrate will result in differing community colonization. The resulting recommendation for monitoring and assessment using cobble baskets in lakes is that baskets be placed on comparable substrate type when comparing between lakes, and that cobble beds be chosen as a more appropriate substrate for deployment, as the added habitat complexity of baskets on soft sediment may act as an attractant and not reflect the true community composition of that habitat.