This study is part of a broader investigation that uses nest boxes to understand the ecological requirements of hollow-dependent wildlife.Summary Habitat preferences need to be understood if species are to be adequately managed or conserved. Habitat preferences are presumed to reflect requirements for food, shelter and breeding, as well as interactions with predators and competitors. However, one or more of these requirements may dominate. Tree-cavity-dependent wildlife species are one example where shelter or breeding site requirements may dominate. We installed 120 nest boxes across 40 sites to target the vulnerable Brush-tailed Phascogale (Phascogale tapoatafa) and the non-threatened Sugar Glider (Petaurus breviceps). The provision of shelter sites where few of quality are available may enable better resolution of habitat preferences. Over three years, we observed the Brush-tailed Phascogale at 17 sites, whereas the Sugar Glider was observed at 39 sites. We tested four broad hypotheses (H1-H4) relating to habitat that may influence occupancy by these species. There was no influence of hollow (cavity) abundance (H1) on either species suggesting our nest boxes had satisfied their shelter requirements. There was no influence of habitat structure (canopy and tree proximity) (H2) immediately around the nest box trees. We found no influence of distance to the forest edge (H3). Variables at and away from the nest box site that appear to reflect foraging substrates (H4) were influential on the Brush-tailed Phascogale. Sugar Glider occupancy was only influenced by a single variable at the nest box site. The lack of influence of any other variables is consistent with the very high occupancy observed, suggesting most of the forest habitat is suitable when shelter sites are available. We found no evidence that the Sugar Glider reduced site use by the Brush-tailed Phascogale.