The aim of the present study was to investigate certain aspects of the reliability and validity of a German version of the Youth Self-Report (total behavior problem score) and to establish norms. In a sample of 252 patients from various child and adolescent psychiatric services and in an unselected sample of 1343 students from nearby schools high coefficients of internal consistency and split-half reliability (> 0.9) were obtained. An additional 105 students were tested and then retested after 5 weeks and again the rank correlations were high (boys: 0.86; girls: 0.90). The patients had higher total problem scores than the non-patients matched for age, sex, socioeconomic status, type of school and status as a foreigner. Patients with a diagnosis on axis 1 (psychiatric syndrome) or axis 5 (abnormal psychosocial situations) of the Multi-axial Classification Scheme were more maladjusted (higher total behavior problem score) than those without such a diagnosis. As expected, the score for the inpatients (N = 99) was higher than for the outpatients (N = 153). Because of the encouraging results, which justify further use of the Youth Self-Report, norms for the total behavior problem score (T-scores, percentiles) were calculated separately for the boys (N = 2085) and girls (N = 2119) from the general population (10-17 years of age).