This research interprets seismicity and geodetic data in terms of the geometry and stressing of locked fault zones which will rupture in large earthquakes. The studies will be carried out in the framework of the asperity model which regards a portion of a seismically active fault zone as being effectively locked against slip, except in large earthquakes, while adjacent portions of the zone slip aseismically, or with lower level seismicity, and do not accumulate comparable stress. The initial focus will be on modelling of the Shumagin Islands subduction segment in Alaska, and the Parkfield region along the San Andreas fault. This research is a component of the National Earthquake Hazard Reduction Program.