The results of observations of auroras and polar magnetic substorms are compared with magnetospheric models. In general, present qualitative or semi‐quantitative magnetospheric models describe fairly well the basic features of auroras and magnetic substorms, or they can probably be adapted to the observations by minor modifications. However, some observations cannot currently be easily interpreted in terms of the models. Of these, the existence of two daytime precipitation zones and some detailed features of the observed particle energy spectrum are among the most important. The stationary models cannot account for the high‐energy part of the auroral particle spectrum seen also in quiescent auroras. Some peaks in the observed spectrum also seem to pose problems. On the whole, the quantitative problems of entry and energization of the auroral particles are still unsol