Aim. The aim of this article is partial empirical verification of the depression image theoretical concept underlying the KIDIOC1 construction, and also, a check of the questionnaire's factor relevancy. Method. KID results of a study of an untreated population sample of 17-year-olds were analysed statistically. Out of 1823 questionnaires, 1349 were included in the analysis (560 filled in by boys and 789 girls by girls). Of these, 499 respondents received a screening diagnosis of depression. 474 sheets were rejected at random to standardise the distribution of the overall scale results. In search of the presence of a general factor and to verify the legitimacy of the division of depressive symptoms according to the clinical criterion, factor analyses were conducted using the principal components method with oblimin, quatrimax and varimax rotations separately and jointly for both sexes. Results. The following new factors were identified: I - pessimism, II - mood instability, III - difficulty in learning, IV - self-destruction, V - fear of the future, VI - eating problems. Conclusions. The analyses conducted only partially confirm the validity of a clinical-picture based questionnaire. A non-compliance of a number of factors with the assumed questionnaire scales emerges. A non-uniform symptomatic depression image in late adolescence phase is confirmed. Two factors stand out decidedly: self-destructive behaviours and eating problems.