Background: Panic attacks are a source of individual suffering and are an independent risk factor for later psychopathology. However, much less is known about risk factors for the development of panic attacks, particularly during adolescence when the incidence of panic attacks increases dramatically. We examined whether internalizing and externalizing problems in childhood predict the onset of panic attacks in adolescence. Method: This study is part of the TRacking Adolescents' Individual Lives Survey (TRAILS), a Dutch longitudinal population cohort study (N = 1,584). Internalizing and Externalizing Problems were collected using the Youth Self-Report (YSR) and the parent-report Child Behavior Checklist (CBCL) at baseline (age 10-12). At age 18-20, DSM-IV defined panic attacks since baseline were assessed with the Composite International Diagnostic Interview (CIDI). We investigated whether early adolescent Internalizing and Externalizing Problems predicted panic attacks between ages 10-20 years, using survival analysis in univariate and multivariate models. Results: There were N = 314 (19.8%) cases who experienced at least one DSM-IV defined panic attack during adolescence and N = 18 (1.2%) who developed panic disorder during adolescence. In univariate analyses, CBCL Total Problems, Internalizing Problems and three of the eight syndrome scales predicted panic attack onset, while on the YSR all broad-band problem scales and each narrow-band syndrome scale predicted panic attack onset. In multivariate analyses, CBCL Social Problems (HR 1.19, p<.05), and YSR Thought Problems (HR 1.15, p<.05) and Social Problems (HR 1.26, p<.01) predicted panic attack onset. Conclusion: Risk indicators of panic attack include the wide range of internalizing and externalizing problems. Yet, when adjusted for co-occurring problem behaviors, Social Problems were the most consistent risk factor for panic attack onsets in adolescence. © 2012 Mathyssek et al.