[en] Two theories of auroral electron acceleration are discussed. The first is the currently widely held view that the acceleration is an ordered process in a quasi-static electric field. It is suggested that, although there are many factors seeming to support this theory, the major qualifications and uncertainties that have been identified combine to cast serious doubt over its validity. The second is a relatively new interpretation in terms of stochastic acceleration in turbulent electric fields. This second theory, which appears to account readily for most known features of the electron distribution function, is considered to provide a more promising approach to this central question in magnetospheric plasma physics. (author)