The many conceptual and methodological difficulties involved in evaluating depression rating scales for children are discussed. A clinical validation of the Depression Self-Rating Scale for Children (DSRSC) is described. The instrument is easy to use and has a predictive value comparable with that of a psychiatric global rating of depressed appearance and history of depression obtained at interview. There was confirmation that the DSRSC can tap an internal dimension of depression and that children are able to evaluate their feeling states. An examination of misclassified children pointed to diagnostic overlap and some unreliability of diagnosis by clinicians.