A statistically determined path of apparent polar wander for the past 300 Ma for North America is given. It has a zigzag form, the bends corresponding to important changes in the drift of North America. Many Mesozoic and Early Tertiary paleopoles from the Western Cordillera do not conform to this path, and they are best explained by motions of miniblocks within the Cordillera. Especially notable is the displacement of Vancouver Island and associated Alaskan terrains (Wrangellia) in the early Mesozoic, and the clockwise rotation of the Coast Range of Oregon and Washington (Siletzia) in the Cenozoic. Limited evidence indicates that part at least of the Northern Appalachian region could have been about 10° south of its present position relative to North America in Late Devonian and Early Carboniferous times, and that it achieved its present position by Late Carboniferous time. Displaced terrains of early and middle Paleozoic age may also be present in the Appalachians. Displacements in the Appalachians and Cordillera could have been caused by the exchange of fragments of continental crust across transcurrent plate junctures. Apparent polar wandering paths for the Precambrian are tentative, but a fairly simple single path can be constructed for all structural provinces of the Laurentian Shield. This reconstruction of a single Precambrian path is supported by agreement among approximately contemporaneous paleopoles from widespread localities. It implies that movements amongst the structural provinces of the shield during the Hudsonian and Grenvillian Orogenies have been modest, and perhaps not in excess of about 1000 km in a latitude sense. This idea has been disputed by those who would apply conventional ideas of plate-tectonics to the Proterozoic, but it has the merit of explaining all the paleopoles and their geological relationships in a comparatively simple unified scheme. Rates of latitude change in the Precambrian may have been over twice as great as in the Phanerozoic.