The constitution of the San Andreas fault zone in central California is modeled by interpreting the recently available geophysical data and by using some broad geochemical and mineralogical constraints. The available geophysical data in this region are consistent with a model that the fault zone throughout the seismogenic depths is composed of saturated, clay-rich fault gouge, with composition similar to the materials exposed along its surficial trace and those recovered from shallow depths. This model appears compatible with the present understanding of mineral stability under the probable PT conditions and the geochemical environment along the fault zone. It implies that materials in the fault zone are the product of extensive hydrothermal alteration of cataclastic rocks with an aqueous fluid rich in CO2 and that large-scale transfer of material between the surrounding crust and the fault zone has taken place. Gravity anomalies and physiographic characteristics of the fault zone provide additional support for this model and suggest that the observed weltlike pressure ridges and uplifted sag ponds along the San Andreas may result from upwelling of low-density fault zone material under the lithostatic pressure still be continuing at the present time. Experimentally determined seismic wave velocities and electrical resistivity of serpentinites in the California Coast Range and elsewhere are mostly significantly higher than the properties of the San Andreas fault zone, suggesting that serpentinite might not be a voluminous fault zone component, although isolated slivers of serpentinite in the fault zone are likely sources for the observed magnetic anomalies.