Self-evaluation procedures are usually used to assess the severity of endogenous depression. In this respect, efficiency tests have hitherto been scarcely considered. A literary survey and our own empirical investigations suggest that self-evaluation procedures should mainly be used in mild and moderate endogenous depressions, and that these procedures give false results in severe and very severe depressions. Efficiency procedures seem not to differentiate well between 'normality' and 'mild depression,' but they become more reliable and valid for depressions of increasing severity. Self-evaluation and efficiency procedures should therefore not be substituted for each other; they should complement each other.