Background and aims: To investigate the impact of obesity on joint pain, we examined the association of body mass index (BMI) with joint pain. Methods: 351 community-dwelling Japanese women aged 40–85 years were asked about joint pain at specific joints, and height and weight were measured. Logistic regression analysis was used to evaluate BMI and other potential predictors of joint pain. Results: BMI was positively associated with knee pain, but not with pain at other specific pints, after adjusting for age. The association of BMI with knee pain remained significant after adjustment for an additional covariate (physical activity). We also examined the associations of age, BMI, and regular physical activity with joint pain at any site, in the arm, in the leg, or in the back. Independently of age and regular physical activity, BMI was positively associated with joint pain in the leg, but not with pain at any site, in the arm or in the back. No significant association of physical activity with joint pain at any site, in the arm, in the leg or in the back was found. Conclusions: Knee pain was associated with greater BMI. This finding supports previous longitudinal studies, suggesting that some knee pain could be prevented by avoidance of excess weight, if the association is causative. However, the association was not very strong; thus, it is likely that many cases of knee pain cannot be avoided by weight reduction only, and may require other interventions.