Seven recording magnetometers monitored time-varying fields at points on a northwest–southeast line 280 km long in north-central Saskatchewan during July 1981. The experiment was designed to test the hypothesis advanced in 1975 by Alabi, Camfield, and Gough that the electrical conductivity anomaly in the North American Central Plains links with the Wollaston Domain in the exposed Precambrian Shield of Saskatchewan. From clear reversals in the phase of vertical variations, it is evident that the conductor passes between two stations straddling the Rottenstone–La Ronge Magmatic Belt, to the immediate east of the Wollaston Domain. Enhanced horizontal variations transverse to the belt at a third, intermediate, station reinforce this interpretation. Vertical-field response arrows obtained from daytime events in the period range 1–40 min clearly indicate the existence of a major conductor that extends to lower crustal depths beneath the belt. To the northwest across the Cree Lake Zone, reversals in the direction of response arrows at short periods (up to 4 min) imply complex electrical structures in the shallow part of the crust.Lewry termed the Rottenstone–La Ronge Belt a Hudsonian "Cordillera-type" arc massif, and described strong geological evidence for collisional suturing and microplate interaction in this part of the Churchill Province. A similar scenario seems to apply in Wyoming, from the work of Hills and Houston. Thus the conductor appears to trace a Proterozoic plate margin 1500 km from a subduction zone in Wyoming along a transform fault to a subduction zone in northern Saskatchewan.