AbstractConsidering several logistical limitations associated with live trapping in arid environments, owl pellet analysis provides a potentially useful alternative to estimate small mammal diversity. Owl pellet analysis has been used to estimate local small mammal diversity under the assumption that the method can reliably reflect the composition of local small mammal assemblages. However, the validity of owl pellet contents for use in ecological assessments has not been established. Thus, the present study aimed to determine the potential for eastern barn owl (Tyto delicatula) pellet analysis to improve the monitoring of small mammals in arid Australia. We analysed two ecological datasets from Pilungah Reserve on the northern boundary of the Simpson Desert in central Australia to assess their effectiveness for detecting expected small mammal species: 250 owl pellets collected from a known roosting site and 5 years of small mammal records from historical, long‐term trapping surveys. Species composition and detection probability of mammals varied depending on the method used. Owl pellet analysis returned higher species richness overall, while traditional trapping methods indicated a higher probability of detecting certain mammal species (e.g., the wongai ningaui Ningaui ridei, brush‐tailed mulgara Dasycercus blythi and the hairy‐footed dunnart Sminthopsis hirtipes). Importantly, analysis of owl pellets revealed two mammal species that had never been detected at the study site (dusky hopping mouse Notomys fuscus and Carpentarian false antechinus Pseudantechinus cf. mimulus). We suggest that owl pellet analysis could be particularly valuable when used as a precursor or in concert with traditional trapping to monitor and assist in the conservation of small mammal species in arid environments.