The golden-cheeked warbler (Setophaga chrysoparia) and the black-and-white warbler (Mniotilta varia) breed in the deciduous/evergreen woodlands of central Texas. The golden-cheeked warbler is endemic, while for the geographically widespread, black-and-white warbler, these woodlands represent the southwestern terminus of the breeding range. To estimate relative abundance and to compare habitat associations, we estimated occupancy, corrected for probability of detection, for both species using replicated point-count surveys of 36 survey stations at each of six sites within the Balcones Canyonlands Preserve Travis Co. TX. Model selection using an a priori, set of candidate models suggested that the probability of detection differed among warbler species and study sites and both species were influenced by time of season, time of day, and habitat characteristics of the survey station—slope and canopy cover. Naïve occupancy estimates suggested that occupancy of black-and-white warblers was much lower than golden-cheeked warblers. However model-averaged estimates of occupancy corrected for imperfect detection suggested that occupancy of black-and-white warblers averaged across all study sites and survey stations were comparable to golden-cheeked warblers. For both black-and-white and golden-cheeked warblers occupancy was positively influenced by slope and this relationship was modulated by canopy cover. Our study provides an example of how occupancy modeling can change naïve perceptions of the relative abundance of species and be used to predict habitat associations of animals through point-count surveys while accounting for imperfect detection.