There is a need to develop appropriate monitoring methods for cryptic reptile species, such as the endangered New Zealand endemic Otago skink (Oligosoma otagense), that take into account incomplete detectability. A recently developed analytical technique was applied to presence–absence data for Otago skinks to derive robust estimates of both detection probability and site occupancy. The estimated detection probability for Otago skinks was 0.4 (s.e. 0.052), and the estimate of the proportion of the study area occupied by Otago skinks was 0.71 (s.e. 0.11). The data derived from presence–absence surveys in consistently sunny weather, with a standardised scanning and searching technique, provided a good basis for robust estimates of detection probability and of the percentage of area occupied by skinks. This survey technique could be applied to other similarly cryptic reptile species to derive rigorous estimates of site occupancy in order to track changes over time or in response to management interventions. It will be particularly appropriate where precise estimation of absolute abundance is not warranted, or where a simple index of relative abundance may be invalid due to failure of the implicit assumption of constant detectability.