In the northwestern part of the Canadian Shield upper Proterozoic rocks of Banks Island, Brock Inlier, Coppermine area (Rae Group), Minto Arch of Victoria Island, and some small inliers have all been considered as parts of the Amundsen Basin. Formational correlation among these areas has been hindered because different stratigraphic subdivisions have been used by geologists working in different regions and because no individual has previously studied the rocks of the basin as a whole. In general, the Proterozoic rocks of the mainland coastal region have been more finely subdivided than those of Victoria Island. Most of the upper Proterozoic rocks of the mainland regions and all of those at the southern tip of Banks Island are considered to be correlative with the Glenelg Formation of Victoria Island. The Glenelg Formation is divisible into at least four widespread mappable units, each of which has diagnostic rock types, sedimentary structures, and in some cases, stromatolites corresponding closely to those of the mainland areas.The stratigraphic succession in the Shaler Group of the Minto Arch area is remarkably similar to that of the sub-Rapitan Proterozoic sequence of the Mackenzie Mountain region about 500 km to the southwest. Approximate time equivalence is proposed. If this correlation is correct then the 'older' Proterozoic sequence in the northern Cordillera is between 1200 Ma and 700 Ma in age and is therefore Neohelikian or Hadrynian in age (or both). Other corollaries of the proposed correlation are that the Rapitan and younger Proterozoic rocks of the Cordilleran region are not represented in the northwestern part of the Canadian Shield and that the older Proterozoic rocks of the Mackenzie Mountains region are probably younger than at least part of the Belt–Purcell sequence of the Canada–U.S.A. border region.New paleocurrent measurements from sandstones in the western part of the Brock Inlier indicate that dominant sediment transport was to the north-northwest. Limited data from the older Proterozoic sequence in the northern part of the Canadian Cordillera suggest transport to the west. These data are consistent with the existence of a weak positive area to the south and east of the two depocentres with possible continuity of a shallow marine shelf sea around its northern margin.