This habitat model was developed to delineate suitable habitat for coastal cactus wren (Campylorhynchus brunneicapillus) in southern California. A primary purpose of the model is to identify potential restoration sites that may not currently support cactus patches required by wrens, but which are otherwise highly suitable. These are areas that could be planted with cactus to increase wren populations, an important management objective for many land managers. We used the Partitioned Mahalanobis D2 modeling technique to construct alternative models with different combinations of environmental variables. Variables were calculated at each point in the center of a 150 m x 150 m cell in a grid of points across the landscape. Variables reflect various aspects of topography, climate, land use (percent vegetation and urbanization at 150 m and 1 km scales), Normalized Difference Vegetation Index (NDVI), and modeled habitat suitability for coastal prickly pear cactus (Opuntia littoralis) and California sagebrush (Artemisia californica). From compiled cactus wren observation data, we randomly selected a total of 845 spatially precise and non-redundant wren locations to use as a calibration dataset and retained the remaining 338 records to use in validation. We randomly selected 1,000 pseudo-absence points from the study area grid to use with the presence validation points to evaluate model performance. For every model-partition, we calculated Habitat Similarity Index (HSI) predictions for presence and pseudo absence points ranging from Very High (0.75 - 1.00); High (0.50 – 0.74); Low (0.25 - 0.49); and Very Low (0 - 0.24). Suitable habitat is identified as grid cells with HSI greater than or equal to 0.5. We calculated Area Under the Curve (AUC) values from a Receiver Operating Curve (ROC) to determine how well models distinguish between presence and pseudo-absence points. We constructed 11 alternative models and selected a best performing model and partition based upon median HSI calibration and validation values and AUC results. The top performing model (Run 11, Partition 1) has an AUC of 0.95 and a median calibration and validation HSI of 0.72 and 0.75, respectively.