We have developed a method for accurate calculation of the distribution of electron energies in the ionosphere under the influence of an electric field. From this calculation we can predict rates of excitation of radiating states of N2, O2, and O as a function of the strength of a hypothetical applied field. We find that it is unreasonable to expect that electric fields are the source of the normal night airglow and that some of the consequences of fields adequate to generate the airglow are incompatible with observed facts. On the other hand, we present evidence that certain auroral forms (especially the so‐called midlatitude red arcs), in which the red lines 6300–6364 A of atomic oxygen dominate the spectrum, are almost certainly caused by local electric fields, perpendicular to the magnetic field, and of the order of 1 mv/cm in strength. This hypothesis, as well as explaining quantitatively the preferential excitation of low‐lying electronic levels in such auroral forms, is consistent with the observed height profiles of the red emission and with the increased drift velocity and decreased density of electrons that are observed to accompany midlatitude red arcs. A field of the magnitude required could be caused by separation of charge in the magnetosphere due to the solar wind. Because the conductivity parallel to the geomagnetic field is very much greater than that perpendicular to the field at all altitudes above 120 km or so, the electric field would exist everywhere in the ionosphere and magnetosphere above this altitude. This mechanism of generation would be consistent with the observed correlation of the occurrence of red arcs with other geomagnetic phen