The practical use of neurobiological markers is increasingly advocated. Research has identified several potential peripheral biomarkers of antisocial behavior that could be applied in clinical practice with relative ease, specifically markers of the autonomic nervous system (ANS) and neuro-endocrinological markers. Researchers advocate that neurobiological markers be studied in combination with psychological and social factors, as it has become increasingly clear that combinations of these factors influence antisocial behavior. However, certain challenges remain that impede the application of these markers in practice. The aim of the current dissertation was to provide new insights into neurobiological parameters related to (the persistence of) antisocial behavior in youth and their potential applicability for clinical practice in youth (forensic) residential care. The chapters of this thesis are all based on the SCREEN study: a national study in 445 youths (aged 14-24 years) within the clinical practice of youth (forensic) residential care settings in the Netherlands. The study consisted of an initial baseline measurement, and three follow-up measurements. Recidivism data were obtained from official records. Several subsamples of the total SCREEN sample were used in the different chapters of this thesis. In order to gain insight in empathic deficits in juveniles with antisocial behavior, self-reported and physiological empathic reactions to empathy induction were studied in relation to psychopathic traits. It was evaluated whether underlying neurobiological measures can contribute to a better understanding of empathy. Self-reported responses and physiological reactions of the autonomic nervous system (ANS) to empathy eliciting film clips were studied in relation to psychopathic traits (n = 416). Results showed that psychopathic traits – especially callous-unemotional traits - were related to lower reported empathy. Moreover, self-acclaimed non-empathic responders showed less physiological reactivity: they showed less heart rate (HR) withdrawal than empathic responders. This infers that, even in a severely antisocial sample, physiological responses may help identify individuals with severely impaired empathic functioning. The biopsychosocial approach on risk assessment was employed by examining the added value of protective and neurobiological factors on top of standard assessment of risk factors (n = 209). Adding protective factors to psychosocial risk factors significantly improved prediction of general reoffending. Adding heart rate variability (HRV) reactivity significantly improved prediction of violent reoffending. More specifically, a blunted decrease in HRV response. Next, we sought to predict quality of life (QoL) from a biopsychosocial perspective. Multivariable mixed model analysis showed that lower trauma and stronger HRV reactivity best predicted increased QoL. It was tested whether the integration of biopsychosocial data could render clinically relevant subgroups related to recidivism. Neurobiological and psychosocial risk factors and recidivism scores were used to identify subgroups in a subsample of 223 juvenile offenders. Latent class regression analysis yielded three relevant subgroups in relation to reoffending behavior. Within these subgroups, specific neurobiological profiles predicted reoffending type (no reoffending, non-violent reoffending and violent reoffending). Psychological, social and neurobiological factors all contributed to the distinction between subgroups, resulting in 98% allocation accuracy. A prototype ‘tool’, designed to assign (new) individuals to the subgroups with accompanying reoffending risk, correctly classifying 92% of these individuals. Finally, several practical recommendations for successful scientific research in youth forensic settings were presented. Being primarily independent, well embedded and accessible, and above all remaining flexible as a research team proved to be key in the SCREEN study. The recommendations can aid researchers in setting up and successfully performing research in different settings of closed youth care. In the seventh and final chapter, key findings of the studies are discussed along the lines of the four distinct tasks suggested by Popma and Raine (2006): diagnostics, risk assessment, intervention options and evaluation of treatment.