A paleomagnetic study of Paleozoic and Mesozoic crystalline rocks in the northern Ford Ranges of Marie Byrd Land, West Antarctica, has determined a middle Cretaceous (circa 100 Ma) paleomagnetic pole and provided constraints on possible clockwise rotation of these ranges and on the rifting of east Gondwana. The 40Ar/39Ar thermochronology data from the Fosdick Mountains record a period of rapid cooling from ~700 ¿C beginning at ~100 Ma. We relate this to extension, intrusion, and uplift associated with the beginning of rifting between Campbell Plateau and Marie Byrd Land. All rocks from the Fosdick and Chester Mountains are normally polarized. We interpret thermochronology and paleomagnetic data to infer that the region was extensively remagnetized in middle Cretaceous time. Inclinations in samples from the Chester Mountains are less steep than those from the Fosdick Mountains, which we interpret as ~25¿ of south tilting of the Chesters. We interpret cooling age data for the time of magnetization to infer that the tilting began after 105 Ma and ended prior to 103 Ma. We further interpret this as constraining the beginning of extension between the Campbell Plateau and western Marie Byrd Land to the interval 105 to 103 Ma. Virtual geomagnetic poles from samples of Early Carboniferous age granodiorite from the western Phillips Mountains lie on the late Paleozoic apparent polar wander path for Australia transferred to Antarctica. Directions from 29 sites in the central and eastern Phillips and Fosdick Mountains give a Middle Cretaceous paleomagnetic pole at 222.3¿ E, 70.5¿ S (A95 6.1¿, KAPPA 20.0). This pole is indistinguishable from other Middle Cretaceous poles for studies further east in Marie Byrd Land. Combining middle Cretaceous poles determined for three other studies of the Antarctic Peninsula, Thurston Island, and the Ruppert-Hobbs coasts with ours gives a Pacific West Antarctic pole at 215.2¿ E, 73.5¿ S (A95 4.0¿, KAPPA 528.9). This pole is discordant by 5¿ to 10¿ from synthetic mid-Cretaceous East Antarctic reference poles, but the degree of discordance is very sensitive to the choice of the specific reference pole. The lack of native East Antarctic reference poles leaves this analysis inconclusive. Accepting 10¿ of discordance, we favor an interpretation where Pacific West Antarctic crustal domains or microplates have rotated clockwise 40¿ to 90¿ and translated a few degrees away from East Antarctica during Late Cretaceous time. ¿ American Geophysical Union 1996