We have compared a simple model of plasma convection at high latitudes with data obtained from simultaneous measurements made by the incoherent scatter facilities at Chatanika, Alaska and Millstone Hill, Massachusetts in June 1978 during moderately disturbed conditions. The measured horizontal plasma drift velocities were averaged for four days to emphasize gross features of the convection pattern and reduce the effects of substorms. The convection model includes the offset of 11.5 ¿ between the geographic and geomagnetic poles, the tendency of plasma to corotate about the geographic pole, and a constant dawn/dusk magnetospheric electric field mapped to a circle about a center offset by 5 ¿ in the anti-sunward direction from the magnetic pole. The radius of the circle corresponds to 17 ¿ of latitude and the electric potentials are aligned parallel to the noon/midnight meridian within the circle. Equatorward of the circle the potential diminishes radially and varies inversely as the fourth power of sine magnetic co-latitude. A consequence of these two offsets and the sunward alignment of magnetospheic electric field is that our model predicts different diurnal convection patterns when viewed at different longitudes in the geographic frame. The concurrently observed diurnal distributions of horizontal plasma convection velocities are different for Chatanika and Millstone Hill even though the measurements cover approximately the same range of magnetic latitudes. We find there is good agreement between our simple model and the gross features of these two diurnal patterns.