Persons with COPD experience co-occurring dyspnea and pain. Little is known about the relationship between symptom co-occurrence with physical activity (PA) and exercise. Novel diagnostic tools are needed for accurate symptom discrimination. In this secondary analysis, we examined relationships between baseline assessments of pain, dyspnea, objectively measured PA, and exercise capacity in persons with COPD who previously enrolled in three PA studies. Pain was assessed with the bodily pain domain of the Veterans RAND-36 (VR-36), and dyspnea with the modified Medical Research Council (mMRC) scale. Average daily step count was assessed with the Omron HJ-720ITC or FitBit Zip pedometer, and exercise capacity with 6-minute walk test (6MWT). We also conducted a pilot neuroimaging study. Neuroimaging data were acquired on a Siemens 3-Tesla Magnetom Prismafit whole-body scanner. Analysis of variance assessed trends in daily step count and 6MWT distance across categories of co-occurring pain and dyspnea. General linear models examined relationships between cortical thickness and resting state functional connectivity (fc) with symptoms and functional status. In 373 Veterans, 98% were male with mean age 70.5± 8.3 years and FEV1% predicted 59 ± 21%. Compared to those with no co-occurrence of pain and dyspnea, those with co-occurrence walked 1,291–1,444 fewer steps per day and had an 80–85 m lower 6MWT distance. Ten males participated in the pilot neuroimaging study. Predominant findings were that lower cortical thickness and greater fc were associated with higher pain and dyspnea, p<0.05. Greater cortical thickness and lower fc were associated with higher daily step count and 6MWT distance, p<0.05. Regional patterns of associations differed for pain and dyspnea, suggesting that cortical thickness and fc may discriminate symptoms. Co-occurring dyspnea and pain in COPD are associated with significant reductions in PA and exercise capacity. It may be feasible for neuroimaging markers to discriminate between pain and dyspnea.