One of the earliest geophysical studies in the Rio Grande rift involved a geomagnetic induction profile across the southern rift by Schmucker (1964,1970). The interpretation of these data, and later studies by Reitzel, Gough, and Porath, clearly established the presence of an electrical conductivity anomaly along the axis of the rift extending from Colorado to west Texas. This anomaly was originally throught to lie well within the mantle. The earliest magnetotelluric (MT) studies in this region generally corroborated the magnetic variation models, although the conducting mantle appeared to be somewhat shallower than previous geomagnetic studies had indicated. On the other hand, more recent MT interpretations differ substantially from the earlier work in that they suggest that the major anomalous features are likely to be in the crust rather than in the upper mantle. Our purpose here is to reconcile the interpretation of the magnetic variation and MT data bases in order to obtain a new preliminary regional model along a latitude of 32¿N that fits both data sets. Applying constraints from MT interpretations, we argue for the presence of an intracrustal layer at a depth of about 20 km, having a conductance of about 2000 S. Allowing for a small portion of the Z field to be of normal and external origin, the magnetic variation data suggest that the conductor disappears or is greatly attenuated to either side of the rift, but the location of its boundaries are not well resolved by the present data. Neither data set could clarity whether the conductor is underlain by normal mantle. On the other hand, neither the MT data nor the magnetic variation data strictly requires the presence of abnormal mantle material beneath the southeastern Basin and Range, the southwestern Great Plains, and the southern Rio Grande rift.