PURPOSE Risk and predictors of long-term mental health outcomes in survivors of adolescent and young adult (AYA) cancers are poorly characterized. Mental health is consequently neglected in long-term follow-up. METHODS We identified all AYA in Ontario, Canada age 15-21 years when diagnosed with one of six common cancers between 1992-2012 using a population-based database, and compared them with matched controls. Linkage to provincial healthcare data allowed analysis of rates of outpatient (family physician and psychiatrist) visits for psychiatric indications and time to severe psychiatric events (emergency room visit, hospitalization, and suicide). Demographic-, disease-, and treatment-related predictors of adverse outcomes, including treatment setting (adult v pediatric), were examined. RESULTS Among 2,208 survivors and 10,457 matched controls, 5-year survivors experienced higher rates of outpatient mental health visits than controls (671 visits per 1,000 person-years v 506; adjusted rate ratio [RR] 1.3; 95% CI, 1.1 to 1.5; P = .006). Risk of a severe psychiatric episode was also increased among survivors (adjusted hazard ratio [HR], 1.2; 95% CI, 1.1 to 1.4, P = .008). Risk of a psychotic disorder–associated severe event was doubled in survivors (HR, 2.0, 95% CI, 1.3 to 2.4; P = .007) although absolute risk remained low (15-year cumulative incidence 1.7%; 95% CI, 1.0 to 2.7). In multivariable analysis, survivors treated in adult centers experienced substantially higher outpatient visit rates compared with those treated in pediatric settings (RR 1.8; 95% CI, 1.0 to 3.1; P = .04). CONCLUSION Survivors of AYA cancer are at substantially increased risk of adverse mental health outcomes, with those treated in adult centers at particular risk. Although absolute incidence was low, survivors were at increased risk of psychotic disorder–associated severe events. Long-term mental health surveillance is warranted, as is research into effective interventions during or after cancer treatment.