Bats are among the least well-known mammals, particularly in terms of their behavior and activity patterns during the winter. Here, we use passive acoustic monitoring to overcome some of the challenges inherent in surveying cryptic forest bats during the wet season to quantify overwintering behavior for 11 species in California coast redwood forests under varying microclimates. Because different species are active at different forest heights, we also examined the effect of acoustic detector placement (treetop or ground level). Generalized linear mixed models were used to relate acoustic detection probability for 8 species to daytime and nighttime temperature, relative humidity, water vapor pressure, and detector placement. The results indicate that daytime maximum temperature best explained variation in nightly probability of detection, and temperature threshold at which bats were predicted to be detected varied considerably across species. By using more precise species detection methods, we were able to resolve significant differences in activity patterns between Myotis yumanensis and M. californicus, 2 species with similar acoustic signatures that are often lumped together. Myotis californicus was predicted to have a 50% probability of detection at maximum daytime temperature as low as 12.5 °C, whereas M. yumanensis was not predicted to have 50% detection probability until maximum daytime temperature was at least 22 °C, suggesting that M. californicus spends less time in torpor. Also, monitoring at the top of the canopy revealed 4 migratory species to be present in the ecosystem on significantly more monitoring nights than could be observed using conventional ground-based monitoring methods. Improving winter bat survey methods provides evidence that diverse bat species are more active in redwood forests during the winter than previously documented. This finding suggests that coastal forests could provide important winter bat habitat for both resident and migratory species.