Psychotic disorders, such as schizophrenia, typically develop in late adolescence and early adulthood, when teens and young adults are transitioning into the most independent and productive years of their lives. Evidence suggests the neurobiological vulnerability for psychosis is likely already present in infancy and early childhood. Across adolescence, this neural vulnerability is unmasked through interactions with abnormal neuromaturational processes and environmental risk factors, ultimately driving the onset of psychosis is early adulthood. Considering that psychotic disorders typically involve a chronic, often treatment-resistant course, early identification and intervention stand to substantially improve the lives of affected individuals, their families, and society as a whole. Within this context, my program of research focuses on three overarching goals: (1) increasing our ability to predict who is at greatest risk for psychosis to improve early identification and intervention, (2) improving our understanding of developmental risk factors for psychosis to inform prevention efforts, and (3) informing the development of novel treatments. For each of these goals, I will share an illustrative example from my research program and outline how I plan to advance this area over the next 5–10 years. In line with the goal of prediction, I will demonstrate how isolating distinct pathophysiological mechanisms of behavioral psychomotor slowing can be leveraged for predicting symptom progression in individuals at high-risk for psychosis. In line with the goal of prevention, I will present findings from a study that critically tests the leading competing explanations for the relationship between adolescent cannabis use and psychosis risk, demonstrating the importance of accounting for shared vulnerability and reverse causality when modeling cannabis–psychosis risk associations. In line with the goal of informing treatment development, I will introduce a current study testing whether non-invasive cerebellar stimulation can ameliorate computational and neural indices of predictive coding abnormalities associated with hallucinations and delusions in patients with schizophrenia. Together, this work illustrates how refining our etiological understanding of the neurobiological processes contributing to psychosis can improve early intervention, reduce illness progression, and optimize recovery.