Plasma wave and plasma measurements from the Dynamics Explorer 1 (DE 1) spacecraft are used to investigate an intense broadband spectrum of low-frequency, <100 Hz, electric and magnetic noise observed at low altitudes over the auroral zones. This noise is detected by DE 1 on essentially every low-altitude pass over the auroral zone and occurs in regions of low-energy, 100 eV to 10 keV, auroral electron precipitation and field-aligned currents. The electric field is randomly polarized in a plane perpendicular to the static magnetic field. Correlation measurements between the electric and magnetic fields show that the perpendicular (~north-south) electric field fluctuations are closely correlated with the perpendicular (east-west) magnetic field fluctuations and the Poynting flux is directed downward, toward the earth. The total electromagnetic power flow associated with the fluctuations is large, approximately 108 W. Two general interpretations of the low-frequency noise are considered: first, that the noise is produced by static fields imbedded in the ionosphere and, second, that the noise is due to Alfven waves propagating along the auroral field lines. For the static interpretation the ratio of the magnetic to electric field strengths at the base of the ionosphere is determined by the Pedersen conductivity, B/(&mgr;0E)=&Sgr;P, whereas for the Alfven wave interpretation it is determined by the Alfven index of refraction, cB/E=nA. Measurements show that the magnetic to electric field ratio decreases rapidly with increasing height. This height dependence is in strong disagreement with the static model if the magnetic field lines are assumed to be equipotentials (E∥=0). At present, no satisfactory model is available for comparison with the data if an electrostatic potential drop is assumed to exist along the magnetic field (E∥≠0). The Alfven wave model is in good agreement with the general form of the height dependence of the magnetic to electric field ratio but disagrees in certain details. The cB/E radio tends to decrease with increasing frequency and is usually somewhat larger than the computed value of the Alfven index of refraction. Some of these difficulties could be accounted for by reflections at the base of the ionosphere or propagation at large angles to the magnetic field (kinetic Alfven waves). For both the static model and the Alfven wave model the source must be located at highaltitudes, since the average Poynting flux is always directed downward, even at radial distances up to 2 RE.