Serious mental illness – depression, bipolar, and psychotic disorder – is among the leading causes of disability worldwide, disproportionately affecting people of non-European ancestry. On top of the burden posed by its symptoms, mental illness is also associated with comorbid health problems. The two most important ‘comorbidities’ of mental illness, given their driving role in decreasing quality and duration of life, are substance (mis)use and cardiovascular disease. Whether these comorbidities arise due to causal relationships is surprisingly unclear. The causal direction is also uncertain: does mental illness lead to comorbidities, and / or do comorbidities increase the risk of (more severe) mental illness? Through several prestigious fellowships, I have established myself as an expert in epidemiological and genetic causal inference methods. In this ambitious project, I will bring together innovative approaches to unravel the relationships of mental illness with substance (mis)use and cardiovascular disease. My aims are to: 1) assess bidirectional relationships between mental illness and its comorbidities by conducting longitudinal analyses in (multi-ancestry) prospective cohort studies, 2) distinguish bidirectional relationships from shared genetic liability by jointly modelling the complete genetic architecture of mental illness and its comorbidities (genomic structural equation modelling), 3) establish causality by using only highly significant genetic variants as instruments for one variable and testing causal effects on another (Mendelian randomization), 4) fully unravel the nature of relationships between mental illness and its comorbidities by ‘triangulating’ evidence from aims 1 to 3, and 5) assess how informing medical doctors about the outcomes of aim 4 influences their clinical decisions in a randomized online experiment. This interdisciplinary project sets the stage for more effective prevention and treatment of mental illness, across ancestry groups.