Persistent pain remains a major challenge in inflammatory arthritis, even when joint inflammation is well controlled. Pain and associated symptoms such as fatigue cannot be explained by peripheral inflammation alone but reflect altered central pain processing. These changes may arise through "top-down" mechanisms, reflecting pre-existing dysfunction in pain perception, or "bottom-up" pathways, driven by peripheral inflammation acting on the brain. Neuroimaging has transformed understanding of these processes by providing in vivo markers of how brain function and structure are related to pain. Functional magnetic resonance imaging (MRI) demonstrates that both task-evoked and resting-state activity are altered in inflammatory arthritis. Connectivity changes involving the thalamus, insula, medial prefrontal cortex, and default mode and salience networks correlate with pain, fatigue, and affective symptoms. Notably, tumor necrosis factor α (TNF-α) inhibitors rapidly normalize pain-related activation, preceding improvements in joint swelling, strongly supporting a bottom-up role for peripheral inflammation. Recent randomized controlled trial data show that baseline central nervous system pain activation predicts analgesic response to TNF-α blockade, positioning neuroimaging as a potential tool for treatment stratification. Complementary modalities provide further insights. Proton electron tomography studies suggest altered pain responses, and novel tracers may clarify contributions of neuroinflammation. Magnetic resonance spectroscopy reveals neurochemical correlates such as increased choline and myo-inositol linked to fatigue, although group-level evidence for overt neuroinflammation remains limited. Structural MRI highlights gray matter changes in regions mediating sensory, cognitive, and affective processing. Together, this supports a dual top-down and bottom-up model of persistent pain in inflammatory arthritis, with important implications for mechanism-based therapies targeting both immune and brain pathways.