We investigated the relationships among motor function, physical activity, and the characteristics of chronic pain (the number of pain sites, pain intensity, and pain-type). Cross-sectional study. An ongoing community-based prospective study conducted in Itoshima, Japan. Community-dwelling Japanese aged 65-75 years (n = 805; 401 men, 404 women). Chronic pain subtypes were examined in terms of the number of pain sites, pain intensity, and pain type. Motor function was evaluated by handgrip strength, walking speed, and the 5 Times Stand-up and Sit Test (FTSST). Locomotive activity, non-locomotive activity, and sedentary time were evaluated by a tri-axial accelerometer as physical-activity parameters. Multiple regression model adjusting for age, sex, education level, employment status, subjective economic status, body mass index, cognitive function, comorbidity, current tobacco use, current alcohol consumption, and regular exercise. In a multivariate analysis, the subjects' walking speed was negatively associated with multisite, moderate-to-severe, and neuropathic-like pain. The FTSST was positively associated with single-site, moderate-to-severe, and neuropathic-like pain. There was no significant association between handgrip strength and any chronic pain subtypes. Locomotive activity was negatively related to multisite, moderate-to-severe, and neuropathic-like pain, but there was no clear association between the amount of non-locomotive activity, sedentary time, and chronic pain subtypes. Severe chronic pain was associated with decreased locomotion-related motor function and physical activity.