I describe a strategy for modeling a specific large scale topographic feature that recognizes the discrepancy between tectonic and geomorphic scales and allows for important feedbacks between these processes. In the geomorphic model, the largest channel in each 4¿4 km cell is explicitly treated, the channel incision being driven by a stream power rule. Hillslopes internal to each cell respond rapidly to incision of the local channel, prompting their treatment in the model as steady state forms. I present analytic expressions for the local relief, which is controlled by the local channel incision rate, and rate constants associated with the dominant hillslope process, either diffusive or landsliding. Initial conditions include any preexisting topography, and a lightly etched a channel network. The model is illustrated using the Santa Cruz Mountains, California. I show that an earlier hypothesis (that the Santa Cruz Mountains are due to advection of topography past an uplift source related to a restraining bend in the San Andreas Fault) remains viable when more specific geomorphic processes are considered. The local tectonics are represented by crustal thickening necessitated by conservation of mass in the bend, right-lateral translation of crust with respect to the bend, and flexural response to the distributed load. The model produces reasonable geographic and statistical distributions of topography using rates of tectonic and geomorphic processes that are within the range of those measured locally. This argues that this restraining bend is a long-lived feature of the San Andreas Fault. The slightly higher crest and shorter length of the southern Santa Cruz Mountains may potentially be explained by slightly faster slip of the crust west of the fault with respect to the bend than of that to the east. Deflection is minimal beneath these relatively small mountains, disallowing the interesting feedback of erosionally driven uplift. The predicted pattern of exhumation could be used to guide the search for fission track sites that could aid in the testing of the model. ¿ American Geophysical Union 1994