BACKGROUND Few studies have focused on the differences between two depressed groups of patients in child psychiatry: the suicidal and the non-suicidal adolescent population. As in other countries, depression is one of the most prevalent diagnoses in adolescents in Hungary. OBJECTIVE The present study was designed to determine (a) whether there are specific symptoms to differentiate between two clinical samples of depressed children: patients expressing suicidal behaviour and their peers with no suicide attempts, and (b) if there are significant differences between parents' and adolescents' reports of the same symptoms. METHODS Using a recently developed semi-structured interview (Diagnostic Evaluation Schedule for Children and Adolescents-Hungarian version, ), 132 symptoms were assessed for two clinical groups of depressed adolescents: a suicidal group (N=54) [corrected], and a non-suicidal group (N=78) [corrected]. The suicidal group had all made an unsuccessful suicide attempt and/or had had frequent suicidal thoughts during the 6 months prior to the study. The non-suicidal group had neither attempted suicide, nor had had suicidal thoughts during the previous 6 months. All cases were selected from a larger sample of 490 consecutively referred new outpatient children over a 1 year period in five psychiatric facilities in Hungary. Only 13-17-year-old adolescents participated in the study. Both samples were identified using operationalised computer algorithm criteria of DSM-IV major depressive disorder episode (MDD) irrespective of the current clinical diagnosis of the patients. The Pearson Chi-square test with Monte Carlo correction was used to evaluate the differences between the suicidal and the non-suicidal depressed samples. RESULTS Hopelessness, negative self-esteem and violent behaviour were the only significant discriminators between the two study groups according to the parent interviews, with increased problem scores in the suicidal sample compared to the non-suicidal sample. Suicidal depressed adolescents view themselves as more depressed and violent than do non-suicidal depressed individuals and were less anxious about their parents. CONCLUSIONS The two depressed samples (suicidal vs. non-suicidal individuals) have only very few dissimilarities. There are, however, some essential differences between the parental and adolescent perceptions of the suicidal and depressive symptoms of the adolescent. The findings of the study underscore the necessity of collecting data from both the parent and the adolescent. CONCLUSIONS Cross-sectional, no lifetime psychopathology, referred samples, no blind estimation of the suicidal status of patients.