The Denali fault system (DFS) extends for ~1200 km, from southeast to south central Alaska. The DFS has been generally regarded as a right-lateral strike-slip fault, along which post late Mesozoic offsets of up to 400 km have been suggested. The offset history of the DFS is relatively unconstrained, particularly at its western end. For this study we calculated relative motion vectors at discrete points along the length of the DFS, based on the well-understood kinematic interaction between the North American, Pacific, and Kula plates, and the following assumptions: (1) The arcuate geometry of the DFS has existed essentially unchanged since the Late Cretaceous; (2) the Yukon-Tanana terrane and other terranes north of the DFS were fixed, in situ, prior to the accretion of the southern Alaskan terranes; and (3) tangential and normal relative motion vector components calculated for points along the DFS using the plate model of Kelley (1993) describe the plate kinematics of the DFS since the Late Cretaceous. The consequent kinematic model for the DFS predicts that left-lateral stresses have acted upon the western end of the DFS for much of its history, and conflicting senses of shear exist between the eastern and western ends of the system. The offset history of the western end of the Denali fault system should be significantly different than the history of the central and eastern sections; consequently, individual crustal blocks in southeast and southwest Alaska may have undergone, respectively, clockwise and counterclockwise rotations. The sense of rotation predicted by our model is in agreement with rotations determined by paleomagnetic studies and provides an alternative model to the ''Alaskan orocline'' hypotheses. ¿ American Geophysical Union 1993