Abstract: Reliable population estimates are needed for the conservation management of seabird populations. Failing to account for detection probability in surveys often leads to underestimate population size and, if detection probability varies among surveys, to bias the estimated trends. This is particularly relevant for storm petrels, which are widespread small burrow- or cavity-nesting seabirds, which have low detection probabilities on land and at sea and whose population status and trends are the least known among seabirds. Here, we used the distance sampling method to estimate detection probability and breeding population size of the cavity-nesting Wilson’s Storm Petrel (Oceanites oceanicus) in the Pointe Géologie archipelago, East Antarctica. Detection probability was 0.353 ± 0.053 and the average density of active nests was 45.53 ± 15.63 nests/ha. The proportion of nests occupied by breeders was estimated using an endoscope on a sample of nests and was 0.455 ± 0.053. The breeding population was estimated to be 793 (95% CI 344–1359) breeding pairs in January 2016. We advocate the distance sampling method as a robust approach to estimate abundance of breeding Wilson’s Storm Petrels in Antarctica. Comparison with an earlier survey suggests that the population has decreased over the past 30 years, possibly partly due to a reduction in nesting habitat following the extension of the surface area occupied by penguin colonies.