Purpose Cancer is the leading cause of nonaccidental death among adolescents and young adults (AYAs). High-intensity end-of-life care is expensive and may not be consistent with patient goals. However, the intensity of end-of-life care forAYAdecedents with cancer-especially the effect of care received at specialty versus nonspecialty centers-remains understudied. Methods We conducted a retrospective, population-based analysis with the California administrative discharge database that is linked to death certificates. The cohort included Californians age 15 to 39 years who died between 2000 and 2011 with cancer. Intense end-of-life interventions included readmission, admission to an intensive care unit, intubation in the last month of life, and in-hospital death. Specialty centers were defined as Children's Oncology Group centers and National Cancer Institute-designated comprehensive cancer centers. Results Of the 12, 938 AYA cancer decedents, 59% received at least one intense end-of-life care intervention, and30%received twoor more. Patients treated at nonspecialty centers were more likely than those at specialty-care centers to receive two or more intense interventions (odds ratio [OR], 1.46; 95% CI, 1.32 to 1.62). Sociodemographic and clinical factors associated with twoormoreintense interventions included minority race/ethnicity (Black [OR, 1.35, 95% CI, 1.17 to 1.56]; Hispanic [OR, 1.24; 95% CI, 1.12 to 1.36]; non-Hispanic white: reference), younger age (15 to 21 years [OR, 1.36; 95% CI, 1.19 to 1.56; 22 to 29 years [OR, 1.26; 95% CI, 1.14 to 1.39]; 30 years: reference), and hematologic malignancies (OR, 1.53; 95% CI, 1.41 to 1.66; solid tumors: reference). Conclusion Thirty percent of AYA cancer decedents received two or more high-intensity end-of-life interventions. In addition to sociodemographic and clinical characteristics, hospitalization in a nonspecialty center was associated with high-intensity end-of-life care. Additional research is needed to determine if these disparities are consistent with patient preference.