The golden-cheeked warbler (Dendroica chrysoparia) is a federally endangered, Neotropical migrant songbird that breeds exclusively in central Texas. Previous studies have identified habitat characteristics associated with the warbler, but a predictive model incorporating data collected over the entire breeding range is lacking. Using logistic regression, we constructed models at 2 scales based on vegetational and topographical data from individual survey points (local-scale) and potential habitat patches (landscape-scale). Models identified similar characteristics important in predicting golden-cheeked warbler occupancy, despite a difference in scale. The models suggested that warbler occupancy was positively associated with steep slopes, forest interior, Ashe juniper (Juniperus ashei) canopy cover, a tall canopy, and proximity to protected lands.