Correlative and process-based approaches to describing the ecological niche in a spatially explicit fashion have often been compared in an adversarial framework. We sought to compare niche models developed via classic (correlative only), niche (process-based information), and hybridized (correlative augmented with process-based derived information) approaches, with the goal of determining if the added effort of process-based model development yielded better model fit. Correlative data layers (i.e., habitat models) included vegetation community types, Euclidean distance statistics, neighborhood analyses, and topographically-derived information. Mechanistic data layers were estimates of thermal suitability derived from field-collected datasets and biophysical calculations, and estimates of prey biomass interpolated from monitoring stations. We applied these models at high resolution (1 m × 1 m pixel size) to habitat occupied by a population of Texas horned lizards (Phrynosoma cornutum) located in central Oklahoma. Results suggested that our treatment of process-based information offered dramatically better identification of suitable habitat when compared to correlative information, but that these results were likely due to low variability of niche variable pixel values. Niche layers nearly perfectly predicted lizard locations; the interpretation of these results suggest that lizards occupy habitat based on thermal suitability over the duration of a field season. Given the low variability observed in thermal suitability layers, we question the ecological reality of these predictions. Correlative models may accurately describe the niche at small spatial scales, and may suffice in situations where time and financial resources are limiting constraints on project goals. Process-based information continues to be an important part of the niche, and may offer additional predictive accuracy via correlative approaches when included in an ecologically meaningful context.