The Calaveras fault is a major component of the San Andreas fault system in the San Francisco Bay area, that generated 13 earthquakes of ML > 5 since 1850. In most recent ML > 5 events, premain shock and postmain shock microseismicity is sparse in the region of coseismic slip. These aseismic areas are believed to represent locked patches of the fault that are accumulating strain to be released in ML > 5 events. We analyze geodetic data to better characterize the spatial distribution of interseismic slip rates on the Calaveras fault, modeling the slip distribution in the seismogenic zone by inversion of over 25 years of surface deformation data. We use a regional fault model with the seismogenic zone of the Calaveras fault discretized into ~6 km ¿ 3 km elements, employing a weighted least squares approach with smoothing and positivity constraints. Our discretized fault slip model consistently identifies regions of slip deficit in the seismogenic zone of the Calaveras fault that generally correspond to regions of decreased microseismicity and ruptures of previous moderate earthquakes. In particular, we find correspondence with the 1979 Coyote Lake and 1984 Morgan Hill events, as well as regions where historical earthquakes on the Coyote and the Sunol-San Ramon segments have occurred. Moment magnitude calculations based on the estimated slip deficit, fault area, and recurrence intervals agree with measured magnitudes of modern events and interpreted historical magnitudes. The results suggest that a combination of geodetically derived fault slip models and microseismicity distribution can be used to characterize seismic hazard.