Between 24-40% of cancer patients in the U.S. use cannabis, principally to manage pain, anxiety, and insomnia. Importantly, evidence suggests that some patients may be substituting cannabis as a strategy to reduce opioid consumption. However, cannabis’ historical classification as a Schedule I substance (i.e., a controlled substance with no known medical use) by the United States (U.S.) Food and Drug Administration has created barriers to conducting rigorous research on its role in cancer symptom management. As a result, the 2021 National Comprehensive Cancer Network’s Adult Cancer Pain guidelines state: “Data supporting the use of cannabinoids as adjuvant analgesics for treatment of cancer pain are extremely limited and the results from what data exist are somewhat conflicting”. Moreover, despite abundant evidence of racial disparities in cancer pain treatment, little is known about the role of cannabis in mitigating racial disparities in cancer pain outcomes. Thus, there is a critical need to conduct rigorously designed research to generate new knowledge of this phenomenon. Using ecological momentary assessment (EMA) methodology and a 12-month prospective cohort design, we propose a multisite study at three health systems in the northeastern U.S. We will enroll 600 (200 per site) ambulatory patients with non-skin solid malignancies who are receiving opioid therapy: 300 cannabis users (weekly use in any form in the prior month) and 300 cannabis non-users (no use in the past 3 months). Of these, 50% will be self-identified African American patients and 50% will be White. Cannabis and opioid use will be assessed via EMA (collected for 1 week/month; 84 days total) and monthly phone surveys and patient reported outcomes (PROs) will be assessed via monthly phone surveys. This study will also describe the poorly understood phenomenon of cannabis use patterns over time by elucidating dynamic within- and between-subject changes in cannabis use, PROs, and opioid use over the course of one year. The Specific Aims are to: (1) describe dynamic within- and between-subject changes in cannabis use over time including frequency, route, source (medical vs. non-medical), indication, and composition; (2) assess if cannabis use over time is associated with key PROs (pain severity and pain-related function, sleep, anxiety, and quality of life) and opioid use (subjective and objective indices) among cancer patients; (3) test if cannabis use moderates the association between race and pain severity; and (4) explore potential moderators of the relationship among cannabis use, PROs, and opioid use including cannabis frequency, source, route, and composition, and current opioid misuse measure. This timely and comprehensive study has high potential to generate new knowledge upon which clinical practice and guidelines related to cannabis use in cancer pain and symptom management may be based. The strong multidisciplinary research team brings the requisite expertise in cancer-related pain, medical cannabis, longitudinal opioid use among patients with cancer, substance misuse, and health disparities. Overall, this research can have a sustained impact on the science of cancer pain and symptom management.