Summary. Magnetotelluric soundings have been made at 25 stations in the Rocky Mountain Trench (RMT) and Main Ranges near 53N, close to the centre of a major conductivity anomaly which had been mapped in a magnetovariation array study. Most stations covered the frequency range 0.01-500 Hz and three stations 0.0002-500 Hz. The resistivity tensor shows low to moderate anisotropy in the RMT, but is strongly 2-D or 3-D in the Rocky Mountains. Apparent resistivities as a function of frequency are displayed in pseudosections along the Trench and along a transverse profile across the RMT and into the Main Ranges. In preparation for 2-D modelling, 1-D inversions have been used to construct resistivity-depth sections satisfying both magnitudes and phases of the MT responses. These show very low resistivities, in the range 1-10S2m, in the upper crust under the RMT and even lower values under the Main Ranges. The latter values give strong confirmation of the Northern Rockies conductor reported by Bingham, Gough & Ingham and are in agreement with models of the conductors fitted to long-period magnetovariation fields by Ingham, Gough & Parkinson. The MT results here reported add some essential depth and resistivity information. It is suggested- that the conductors beneath the Rocky Mountains Main Ranges and Trench constitute a thickening at the edge of the Canadian Cordilleran Regional (CCR) conductor. Gough has argued that a wide variety of geophysical and geological parameters indicate high temperatures and partial melting in the mantle under the CCR conductor. At the upper crustal depths penetrated in this magnetotelluric study, it is considered more probable that the high conductivity is caused by hot, saline water of mantle origin rather than silicate melt. The CCR in general may have two layers of fluid producing its high conductivity, silicate melt below and saline hot water above.