[Correction Notice: An Erratum for this article was reported in Vol 24(6) of European Child & Adolescent Psychiatry (see record 2015-25233-004). Unfortunately, the names of two authors, Ulrike Ravens- Sieberer and Fionna Klasen, were omitted in the original publication of the article. The corrections are present in the erratum.] Background: Scientifically sound and valid information concerning course and prediction of mental health problems in children and adolescents in the general population is scarce, although needed for public mental health issues and daily clinical practice.Objectives: The psychopathological profiles of children and adolescents were analysed using the parent version of the Strengths and Difficulties Questionnaire (SDQ-P) in a longitudinal setting, also investigating the predictive value of the SDQ-scores.Methods: SDQ’s total psychopathological difficulties, emotional symptoms and hyperactivity-inattention scores of n = 630 children and adolescents (age 6-18;11 years) were examined along four assessment measurement points (T0-T3) over 6 years, using data from the BELLA study. According to the English normative data, the participants were categorized as ‘normal’, ‘borderline’ or ‘abnormal’ based on their SDQ-scores. Groups remaining within categories were descriptively determined by means of frequency analysis, a subsequent graphical evaluation displayed the transitions from T0 to T3 concerning the different categorical classifications. Finally, ordered probit regression was used to examine whether age, gender, socio-economic status (SES) and baseline impact-score (IS) correspond to the SDQ-predicted classification.Results: As expected, low SES and high SDQ-IS were associated with significantly increased scores on all examined SDQ-scales. Regarding the long-term aspect of SDQ-scores it could be shown that most children and adolescents remained ‘normal’ over a measurement period of 6 years, while only a small number of children and adolescents steadily remained ‘abnormal’ or newly developed mental health problems, respectively. For example, on the ‘hyperactivity-inattention’-scale, only 1 % of the children and adolescents changed from ‘normal’ to ‘abnormal’ (T0-T3), whereas on the ’emotional symptoms’-scale, 7 % changed from ‘normal’ to ‘abnormal’ (T0-T3). In general, the SDQ-category ‘borderline’ and specifically the subscale ’emotional symptoms’ change in both directions. Abnormal SDQ-scores at baseline, SES, gender and IS were related to the prediction of the SDQ-sores at T3.Conclusion: An SDQ-screening of children and adolescents may help for early detection, prediction and treatment planning. Also, these results may contribute to a better understanding of the course of mental health problems in childhood and concurrently may allow a better psychoeducation and prevention.