A simple model of collisionless plasma kinetic theory [Knight, 1973> has been used to predict the relationship between the upward parallel current and the parallel potential drop. The DE 1/DE 2 pair offers a unique opportunity to test this relationship because the DE 1 spacecraft can measure the high altitude plasma parameters without contamination from auroral heating. The field-alighed current density from Knight's formula is almost linear in the ratio of j∥/e&PHgr;∥ for values of e&PHgr;∥/kT between 1 and 10. We find, using measured values of J∥ (mapped to the surface) and &PHgr;∥ (inferred from measurements either at high altitudes or low altitudes), the ratio of J∥/e&PHgr;∥ varies considerably but with a mean value about 0.5~2.2¿10-9 mho/m2. Using either the full Knight's formula or the linear simplification, we can predict the parallel potential drop (given the high-altitude density, temperature, current, and magnetic mirror factor) to test against the measured potential drop, or we can predict the current given the potential drop and the high-altitude plasma parameters. The agreement is generally quite good and the linear simplification is reasonable; however, the Knight current, since it does not include upward ionospheric electrons, is not valid in the downward return current region. We also show that suprathermal electron bursts are observed in the diffuse aurora at the same invariant latitudes, both at high and at low altitudes. Thus we suggest that these ''bursts'' are more properly described as a spatial rather than temporal phenomenon. ¿ American Geophysical Union 1991