Ionospheric convection inferred from Super Dual Auroral Radar Network (SuperDARN) HF radar measurements is compared with an equivalent ionospheric convection derived from ground magnetometer data in the dayside winter high‐latitude ionosphere. Although there was general agreement between observed convection patterns produced by radars and magnetometers, there were significant differences in details. The orientation of equivalent convection vectors inferred from magnetic data was often opposite to the convection vectors determined by the SuperDARN radars in the poleward part of the convection vortex structure, though the agreement was reasonable in its equatorward part. The magnitudes of convection vectors determined from radar data and those inferred from magnetometer data were often different. The observed differences are attributed to strong horizontal inhomogeneity in the ionospheric conductivity distribution for winter conditions. It is possible that magnetic disturbances in the dark high‐latitude ionosphere are strongly affected by field‐aligned currents at the terminator that separates regions of the sunlit highly conducting ionosphere and dark poorly conducting ionosphere.