The onset timing of the NAM can lengthen or shorten the fire season in the southwestern United States. NAM onset dates for the southwestern United States and each Predictive Service Area (PSA) are determined from 1992-2011. Thresholds of onsets and busy periods of fire activity were determined by PSA and for the entire southwestern United States. K-means clustering and Principal Component Analysis (PCA) were implemented to identify atmospheric patterns and centers of action contributing to the NAM onset and busy periods for each PSA and the entire southwestern Untied States. Some of these patterns correspond to known atmospheric patterns that create intraseasonal precipitation variability for the NAM including critical fire weather patterns for the region. Relevant surface and atmospheric variables including fuel moisture and lightning were utilized in classification and regression tree (CART) and boosted regression tree (BRT) analyses to develop a model for fire and large fire occurrence by PSA. The CART and BRT models provide daily predictions for fire occurrence that can be enhanced by the identification of busy fire activity and NAM onset atmospheric patterns to provide improved and comprehensive fire potential forecasts for the southwestern United States.