Estimating the distributions of cryptic species is essential for conservation, yet our understanding is hampered by animal behavior and imperfect detection. We developed and implemented a multiscale occupancy survey protocol to estimate the probability of occurrence, probability of being active on the surface, and detection probability of two range-restricted terrestrial salamanders, Scott Bar Salamanders (Plethodon asupak) and Siskiyou Mountains Salamanders (P. stormi), in interior northern California, USA. We established survey sites near locations of historical occurrence of these salamanders and surveyed each site on one to six visits in late fall 2023 and spring 2024. We compared models with different environmental variables for predicting salamander occurrence and surface activity. Much model selection uncertainty in the effects of covariates existed, but the model with most support indicated that Plethodon salamanders in interior northern California were more likely to occur at sites near recent historical occurrences that had longer growing seasons after controlling for elevation and aspect. Plethodon stormi or P. asupak surface activity was higher at night than during the day and was highest when substrates were cool and moist. Our survey protocol was successful for quantifying the effects of site and visit characteristics on P. stormi or P. asupak occurrence and availability; minor modifications will likely improve its utility for providing unbiased inference about salamander occurrence and surface activity.