The northern long-eared bat (Myotis septentrionalis) is currently listed as threatened under the U.S. Endangered Species Act largely due to population declines resulting from the spread of white-nose syndrome in North America. White-nose syndrome was confirmed in Iowa in 2015, emphasizing a need to closely monitor populations of M. septentrionalis statewide. We applied presence-only models to predict landscape-scale resource selection by M. septentrionalis using roost tree observations and mist net captures from various research and environmental assessment projects in Iowa (2003–2015). We used a simultaneous autoregressive (SAR) model to account for residual spatial autocorrelation in our compiled data set and estimate the proportional probability of use of summer habitats for M. septentrionalis. We estimated SAR models using four environmental predictor variables measured at two landscape scales (0.5- and 2.4-km) representative of M. septentrionalis home range sizes in North America. The SAR models resulted in high predictive fit with withheld test observations and an independent data set of acoustic detections of M. septentrionalis from recent surveys (2016–2018), indicating a significant positive relationship existed between habitat quality (as an index of selection) and distribution of M. septentrionalis at landscape scales. At both spatial scales, M. septentrionalis showed positive selection of closed canopy interior forest, bottomland hardwood forest, and total perennial stream length, whereas at the 0.5-km scale, M. septentrionalis also showed a positive association with open canopy forest. Our models indicated that up to 7.0% and 8.5% of the state was comprised of potentially suitable forested summer habitats for M. septentrionalis for 0.5- and 2.4-km scales, respectively. Our models also indicated the distribution of highly selected habitats at landscape scales in Iowa and accurately predicted independent observations of M. septentrionalis in areas of the state where no capture efforts have occurred. This study provides methods to predict landscape-scale resource selection and distribution for bats where multiple fine-scale data sources exist across broad geographic regions.