H. G. Booker's (1956) theory of the radio aurora as a random assembly of field-aligned irregularities has been unsuccessful in explaining many observations, and this lack of success has led to the application of plasma instability theory to the phenomenon. The ion-acoustic instability theory of D. T. Farley (1963) has had some success, and evidence is rapidly accumulating that the ion-acoustic instability is an important mechanism in the production of radio aurora. A second theory, originally developed by A. Simon (1963), is currently being applied to the radio aurora. This theory involves another type of plasma instability, which is here called the ‘drift-gradient instability.’ Both instabilities require the presence of an electric field directed perpendicular to the ambient magnetic field. For the ion-acoustic instability the critical condition is that the electric field exceeds a threshold value of the order of 30 mv m−1. The drift-gradient instability, however, requires a positive gradient of ionization density in the direction of the electric field. The growth of this drift-gradient instability is dependent on the condition that electron-ion drift velocity divided by the scale length of the plasma density exceeds a value such that damping due to diffusion is overcome. Experimental data are reviewed in the light of the two theories. It is shown that all the major features of the so-called diffuse radio aurora (type B1 in the provisional nomenclature of the International Association of Geomagnetism and Aeronomy, IAGA, 1968) can be explained by the ion-acoustic instability theory if account is taken of the nonlinear production of secondary irregularities discussed by J. P. Dougherty and D. T. Farley (1967). Progress to date in applying the drift-gradient instability theory suggests that most features of the discrete radio aurora (type B2 and B3 in the IAGA classification) may be explained. No attempt is made at this stage to relate the theories to forward scatter observations of the radio aurora. The theories suggest that appropriate observation of the radio aurora may be used as a powerful ground-based tool for quantitative studies of electric fields in the ionosphere. Existing experimental data are discussed from this point of view, attention being paid in particular to substorms and pulsations of the B1 type radio aurora in the Pc 5 frequency range. Results suggest that Pc 5 magnetic pulsations are associated with standing hydromagnetic waves in the magnetosphere with, contrary to some currently held assumptions, an antinode of electric field and field line displacement in the ionosphere. Progress in relating B1-type radio auroral events to IPDP micropulsation events is reviewed. If a relation can be established, it will allow further deductions to be made concerning the electric field in the magnetosphere. The aspects of the ionospheric electric field that can be monitored by an auroral radar system and the specifications for such a system are outlined.