Remotely sensed imagery, coupled with wildlife habitat models provide a powerful tool for the implementation, assessment, and monitoring of wildlife conservation/restoration initiatives. Observed, empirical relationships between a species abundance metric and landscape structure/composition are used to structure models. Habitat suitability models always represent a trade off between breadth of applicability and specificity. Large-spatial extent, coarse spatial resolution data sets may be useful for characterizing potential animal distributions at regional or continental scales; however, habitat models developed at this spatial scale may have little applicability for predicting suitability at finer spatial resolutions. Whereas numerous issues related to multi-scale analysis have been acknowledged with respect to wildlife habitat models, only recently have sources of high-resolution imagery been readily available for site-specific analyses. We outline a multi-scale approach to habitat modeling and demonstrate this approach with northern bobwhite. We developed a coarse resolution model appropriate for identifying focal regions likely to support bobwhite using classified LandSat imagery and relative abundance measures from breeding season call counts. Then we developed a fine resolution model based on 4-m multispectral IKONOS imagery and animal space-use for planning and implementing conservation practices at the local scale. We discuss the application of this hierarchical approach to conservation planning.