Conclusions Analysis of lesser sooty owl pellets is a particularly useful method for compilation of species inventories of small to medium-sized mammals, being more effective than standard live-trapping surveys within the rainforests of the Wet Tropics. Implications Owl-pellet analysis is well suited for monitoring mammal communities, as long as periodic data are collected from the same roost(s) and the pellets have been deposited by the same individual bird(s). Additional research relating to variability in behavioural traits between individual lesser sooty owls that have potential to confound results via sampling bias must be undertaken before owl-pellet data can be used for comparison of mammal community structure among sites, or for monitoring a site over a period spanning a change in the individual owls that are depositing pellets at that site.Context Predictive modelling of the impacts of climate change has highlighted a need for on-ground monitoring of mammal communities within the Wet Tropics bioregion of Queensland to inform wildlife conservation management. Standard mammal-survey techniques are labour intensive and often incapable of detecting rare, trap-shy or otherwise elusive species, effectively rendering them invisible to researchers even when substantial resources are applied. Aims To assess the utility of feeding remains (regurgitated pellets) from lesser sooty owls, Tyto tenebricosa multipunctata, for monitoring populations of small to medium-sized mammals within the Wet Tropics. Methods Three female lesser sooty owls were radio-tracked to determine home-range areas. Radio-tracking and targeted systematic searches were used to locate diurnal roosts containing owl pellets from which prey species were identified. Owl-pellet data were directly compared with live-trap data within three putative owl home ranges. Additionally, analyses of a dataset spanning more than 15 years demonstrated the utility of owl-pellet data for monitoring mammal communities. Key results Owl-pellet surveys yielded 14 mammal species (840 individuals) from 152 man-hours, compared with six mammal species (361 individuals) from 194 man-hours of live trapping. Both survey methods identified Rattus fuscipes, Melomys spp. and Antechinus spp. as the most abundant species but live-trap data were found to under-represent relative abundance of Melomys spp. and over-represent relative abundance of R. fuscipes in comparison to owl-pellet data.