Changes to bat distribution and habitat associations at the local to sub-landscape scale in the post white-nose syndrome (WNS) environment have received little attention to date despite being critical information for managers. To better understand the spatial nature of bat population declines, we modelled both activity patterns and occupancy from acoustic surveys for the Myotis lucifugus (little brown bat) on Fort Drum Military Installation in New York, USA over 15 summers (2003–2017) that span the pre-WNS, WNS-advent (2008) and post-WNS periods, using a set of generalized linear mixed models and geospatial analysis. Our best supported model indicated significant differences between years with significant declines in activity post-WNS. M. lucifugus activity was most closely associated with woody wetland habitats over the study period, however, the spatial patterns of high activity areas were variable over years, with the areal extent of these high activity areas decreasing post-WNS. Our best supported occupancy model varied by year. However, the null occupancy model [Ψ(.)] was either competing (within 2 ΔAIC units) or was the best supported model. Meaning that none of our environmental variables seemed to impact occupancy, and when they did, these differences were not significant. There was high disagreement between our relative activity models and predictions compared to our occupancy models, suggesting that geographic spatial scale and the resolution of the data impacts model outcome. Our results indicate that continued acoustic monitoring of bat species in the Northeast to assess ongoing temporal and spatial changes in habitat associations and to provide direction for future mist-netting studies should rely more on relative activity as the metric of choice rather than site occupancy.