We present our work towards the identification of neurobehavioural markers for externalizing disorders in the longitudinal imaging genetics project IMAGEN of 2000 adolescents (Schumann et al. Mol Psychiatry 2010). This includes studies on the heritability of brain structure (Toro et al. Mol Psychiatry 2015) and function (Dickie et al. PloS Genetics 2014), the relation of gene expression and brain functional activity (Richiardi et al. Science 2015), as well as a neuropsychosocial model of prediction of adolescent alcohol abuse (Whelan et al. Nature, 2014) and discoveries of novel brain mechanisms underlying reinforcement-related behaviour, including impulsiveness (Whelan et al. Nature NS 2012) and reward anticipation (Jia et al. PNAS 2016, in preparation). We will describe a model that enables integration of different modalities while maximizing correlation between biological mechanisms and observable behaviour (Ing et al. Nature NS, in revision). We propose that a combination of neurobehavioural investigations with advanced statistical models will result in the development of robust multimodal biomarker profiles for clinical application. However, as both psychopathology and brain function are influenced by environmental factors (Quinlan et al. Am J Psych 2017) it is necessary for precision medicine to be applicable in a global setting that it takes into account local environmental determinants of mental illness. For this reason we have established and will introduce a consortium on Global Imaging Genetics in Adolescents (GIGA) that brings together samples Europe, India, China and the U.S. to comparatively investigate environmental effects on brain development and behaviour in a global context.