Traditional pathogen surveillance methods for white-nose syndrome (WNS), the most serious threat to hibernating North American bats, focus on fungal presence where large congregations of hibernating bats occur. However, in the western USA, WNS-susceptible bat species rarely assemble in large numbers and known winter roosts are uncommon features. WNS increases arousal frequency and activity of infected bats during hibernation. Our objective was to explore the effectiveness of acoustic monitoring as a surveillance tool for WNS. We propose a non-invasive approach to model pre-WNS baseline activity rates for comparison with future acoustic data after WNS is suspected to occur. We investigated relationships among bat activity, ambient temperatures, and season prior to presence of WNS across forested sites of Montana, USA where WNS was not known to occur. We used acoustic monitors to collect bat activity and ambient temperature data year-round on 41 sites, 2011-2019. We detected a diverse bat community across managed (n = 4) and unmanaged (n = 37) forest sites and recorded over 5.37 million passes from bats, including 13 identified species. Bats were active year-round, but positive associations between average of the nightly temperatures by month and bat activity were strongest in spring and fall. From these data, we developed site-specific prediction models for bat activity to account for seasonal and annual temperature variation prior to known occurrence of WNS. These prediction models can be used to monitor changes in bat activity that may signal potential presence of WNS, such as greater than expected activity in winter, or less than expected activity during summer. We propose this model-based method for future monitoring efforts that could be used to trigger targeted sampling of individual bats or hibernacula for WNS, in areas where traditional disease surveillance approaches are logistically difficult to implement or because of human-wildlife transmission concerns from COVID-19.