Some currently used tree-scale sampling techniques targeting saproxylic insects capture individuals that are attracted to or landing on specific potential hosts. The success of such techniques is entirely dependent on strong primary attraction in targeted insects. However, up to this point, field experiments testing the primary attraction hypothesis have produced contradictory results. To test the efficiency of such techniques, and consequently, the strength of primary attraction for saproxylic Coleoptera, we sampled insects landing on contrasting snag types including new and old snags of five different tree species using sticky traps in a single mixed 135-yr-old boreal stand in Western Quebec, Canada. Ordination analyses showed homogenous assemblages among the different snag types and stovepipe controls, when considering either all species captured or only targeted functional groups, and very few species showed strong affinities to specific snag types. Species composition of assemblages was in several cases correlated with the species and status of trees neighboring the sampling units, which suggest that prelanding host selection mechanisms do not allow insects to single out a potential host while in flight. Our results suggest that primary attraction may play a role at larger spatial scales and help insects identify potential habitat patches, while selection of a single host at the local scale is done by trial-and-error through random landing. In such a context, future studies aiming at describing precise host-use patterns of saproxylic insects should rely on methods targeting larvae or emerging adults such as wood dissection and rearing.