A major obstacle to reliable population assessment and monitoring programs is the fact that the detectability of populations or individuals is usually imperfect, meaning that estimates of area of occupancy or abundance will be biased low unless methods are used which can account for imperfect detection. We surveyed water bodies in three parts of the historical distribution of the Alpine Tree Frog (Litoria verreauxii alpina) in southeastern Australia to determine probabilities of detection for adults and tadpoles. Combined static call surveys and spotlight surveys were used to detect adult frogs, and combined visual surveys and funnel trapping surveys were used for detection of tadpoles. A very high estimated probability of detection (>0.9) was obtained during the tadpole surveys, whereas the estimated probability of detection during surveys for adults was considerably lower (<0.7) and varied between observers. Due to their high probability of detection, we recommend the use of tadpole surveys as a survey and monitoring technique for L. v. alpina.