Abstract Chronic pain due to knee osteoarthritis (OA) is a large contributor to disability, affecting over >14 million Americans. However, treatment is difficult and outcomes are poor. This is because pain is not monolithic, and differences in underlying pain mechanisms affect treatment response. While knee OA pain can be due to tissue and joint damage (nociceptive pain), it can also be augmented and maintained by central nervous system (CNS) dysfunction – i.e., centralized or nociplastic pain. Variable CNS contributions explain why some people have severe radiographic knee damage yet report no pain, while others have “normal” radiographs yet report severe pain. Our group and others have shown that this centralized phenotype occurs in nearly all chronic pain conditions. Further, we have demonstrated the clinical relevance of matching pain mechanisms with therapies, showing that a greater degree of pain centralization lowers responsiveness to interventions (e.g., opioids and surgeries) directed toward reducing nociceptive pain in knee OA. Differences in underlying mechanisms may explain inconsistent clinical trial results with cannabinoids - the active compounds in Cannabis sativa. Clinical trials suggest that cannabinoids can be effective analgesics, but this has been shown primarily with Δ[9]-tetrahydrocannabinol (THC) dominant preparations, which have abuse potential. However, a recent study showed that cannabidiol (CBD) reduced pain and increased function in men with knee OA. CBD is non-intoxicating, and exerts analgesic and anti-inflammatory effects. The proposed studies are the first attempt to understand how CBD and THC affect different chronic pain mechanisms by examining the effects of these compounds on knee OA in individuals with varying degrees of pain centralization. Our overarching hypothesis is that CBD will decrease peripheral inflammation, THC will modify CNS pain processing, and combined CBD+THC will do both. To test this hypothesis, we propose three specific aims. Aim 1: In a randomized, double-blinded, 2x2 factorial design study, longitudinally assess peripheral and CNS effects of CBD and THC on inflammation (interleukin 6) and neurobiological correlates of centralized pain (i.e., default mode network to insula connectivity) using a phenotyping battery of patient reported outcomes, experimental sensitivity testing, and neuroimaging that we have successfully applied to other drugs and pain conditions. Aim 2: Examine effects of THC and CBD and their metabolites on additional indices of brain connectivity and neurochemistry as well as inflammatory markers. Aim 3 (exploratory): Assess whether pain centralization predicts differential analgesic responsiveness to CBD and THC. Given that pain centralization occurs on a spectrum in many chronic pain conditions, our approach in knee OA will act as a translational model that can be applied to nearly all other pain conditions and will have broad implications for developing non-opioid ana...