There is a well-established historical observation that structural joint damage by plain X-ray correlates poorly with symptomatic disease in osteoarthritis (OA). This is often attributed to the inability to visualise soft-tissue pathology within the joint and the recognition of heterogeneous patient factors that drive central pain sensitisation. A major issue is the relative paucity of mechanistic studies in which molecular pathogenesis of pain is interrogated in relation to tissue pathology. Nonetheless, in recent years, three broad approaches have been deployed to attempt to address this: correlative clinical studies of peripheral and central pain outcomes using magnetic resonance imaging, where soft-tissue processes can be visualised; molecular studies on tissue from patients with OA; and careful molecular interrogation of preclinical models of OA across the disease time course. Studies have taken advantage of established clinical molecular targets such as nerve growth factor. Not only is the regulation of nerve growth factor within the joint being used to explore the relationship between tissue pathology and the origins of pain in OA, but it also provides a core model on which other molecules present within the joint can modulate the pain response. In this narrative review, how molecular and pathological tissue change relates to joint pain in OA will be discussed. Finally, a model for how tissue damage may lead to pain over the disease course will be proposed.