Cretaceous and early Tertiary arc and forearc rocks found along the coast in southern California and Baja California have been shown by paleomagnetic measurements to have originated many hundreds, or even several thousands, of kilometers south of their current locations. Northward transport also is found in Cretaceous batholithic rocks near the edge of the continent in Washington and British Columbia. The consistency of this pattern suggests that slices of arc and forearc rock originating on the western edge of North America have been translated along the coast by strike‐slip faulting. Strike‐slip faults are shown to be an expected response to north‐oblique subduction of the Farallon and/or Kula plates. Faulting and subduction probably were concurrent activities; i.e., the arc and forearc grew and were displaced simultaneously. Because displacement was dominantly parallel to subduction‐related lithic belts, arc and forearc rocks may have traveled large distances without producing either significant hiatuses in the rock record or conspicuous disruptions in the distribution of lithic belts. Subduction‐related strike‐slip faulting is shown to be favored by high angles of oblique convergence, shallow dip of the subducting slab, relatively easy slip on potential faults within the arc or forearc, and the existence of a place for the moving sliver to go. It is suggested that these conditions are not difficult to satisfy and that subduction‐related strike‐slip faulting may thus be a common feature of convergent orogens. The location of the San Andreas fault inboard from the continental margin may be attributable to localization of the Pacific‐North American transform by a system of precollision subducti