New geologic mapping and geochronology show that margin-parallel strike-slip faults on the western limb of the southern Alaska orocline have experienced multiple episodes of dextral motion since ~100 Ma. These faults are on the upper plate of a subduction zone ~350-450 km inboard of the paleotrench. In southwestern Alaska, dextral displacement is 134 km on the Denali fault, at least 88-94 km on the Iditarod-Nixon Fork fault, and perhaps tens of kilometers on the Dishna River fault. The strike-slip regime coincided with Late Cretaceous sedimentation and then folding in the Kuskokwim basin, and with episodes of magmatism and mineralization at ~70, ~60, and ~30 Ma. No single driving mechanism can explain all of the ~95 million-year history of strike-slip faulting. Since ~40 Ma, the observed dextral sense of strike slip has run contrary to the sense of subduction obliquity. This may be explained by northward motion of the Pacific plate driving continental margin slivers into and/or around the oroclinal bend. From 44 to 66 Ma, oroclinal rotation, perhaps involving large-scale flexural slip, may have been accompanied by westward escape of crustal blocks along strike-slip faults. However, reconstructions of this period involve unproven assumptions about the identity of the subducting plate, the position of subducting ridges, and the exact timing of oroclinal bending, thus obscuring the driving mechanisms of strike slip. Prior to 66 Ma, oblique subduction is the most plausible driving mechanism for dextral strike slip. Cumulative displacement on all faults of the western limb of the orocline is at least 400 km, about half that on the eastern limb; this discrepancy might be explained by a combination of thrusting and unrecognized strike-slip faulting.