The blue crab, Callinectes sapidus, is an ecologically and economically valuable species in Chesapeake Bay. Field surveys and laboratory experiments indicate that blue crab mortality is significant during severe winters. We applied a temperature and salinity-dependent survival model to empirical temperature and salinity data to explore spatial and interannual patterns in overwintering mortality. Harmonic regression analysis and geostatistical techniques were used to create spatially explicit maps of estimated winter duration, average temperature, average salinity, and resulting crab survival probability for the winters of 1990-2004. Predicted survival was highest in the warmer, saline waters of the lower Bay and decreased with increasing latitude up bay. There was also significant interannual variation with survival being lowest after the severe winters of 1996 and 2003. We combine the survival probability maps with maps of blue crab abundance to show how winter mortality may reduce blue crab abundance prior to the start of the harvesting season.