BACKGROUND: Understanding the etiology of psychosis is essential to the development of preventive interventions. The COVID-19 pandemic provides a rare natural experiment that can expand our understanding of the role of social factors in the trajectories and etiology of psychosis across adolescence, particularly in Tokyo where the prevalence of actual COVID-19 infection remained low. We hypothesized that the likelihood of self-reporting psychotic experiences (PEs) would increase following the onset of the COVID-19 pandemic. METHODS: The Tokyo Teen Cohort (TTC) is a prospective cohort study of adolescents in the general population of the Tokyo metropolitan area, followed from age 10 to 16 years. We used multi-level linear regression models to test the associations between the phase of the COVID-19 pandemic and self-reported PEs. RESULTS: Among 1935 adolescents included in the analysis, a rapid increase in PEs occurred at the onset of the COVID-19 pandemic, following approximately 6 years of steady decline across prior waves. This association was more pronounced for boys compared to girls. This increase became more pronounced as the pandemic moved into later phases, defined based on contemporaneous sociopolitical changes in Tokyo (i.e. changes to school closure, social distancing guidelines, and the state of emergency status). CONCLUSIONS: The steady decline in PEs across adolescence was halted and reversed concurrent with the COVID-19 pandemic onset, despite very low rates of COVID-19 infection. This implicates COVID-19 related socioenvironmental factors as contributory etiological factors in the development of PEs in this adolescent cohort.