OBJECTIVES: To investigate potential factors associated with joint pain among postmenopausal women. DESIGN: Population-based cross-sectional study. SUBJECTS: Six hundred and ninety postmenopausal Japanese±American women from the Hawaii Osteoporosis Study (age: 55±93 y). MEASUREMENTS: Data for this study were collected at the 1992±1994 examination of the Hawaii Osteoporosis Study, except that data on spinal osteoarthritis were obtained based on radiographs at examinations before 1987. Information on painful joints at a variety of skeletal sites, smoking, and physical activity was collected by questionnaire. Bone density was measured at the lumbar spine, distal and proximal radius, and heel. Quantitative bone ultrasound was also measured at the heel. Prevalent vertebral fractures were identi®ed on lateral spine radiographs using 3 s.d. below the normal population mean as the cutoff. Non-spine fractures were identi®ed based on self-report, and were veri®ed using medical records. RESULTS AND CONCLUSIONS: In this cross-sectional analysis, bone density, quantitative bone ultrasound, prevalent vertebral fractures, and non-spine fractures were not signi®cantly associated with joint pain. Greater body weight or body mass index were signi®cantly and positively associated with joint pain at most weight-bearing joints. The results suggest that a substantial proportion of joint pain at these sites could be prevented by avoidance of excess weight, provided the association is causative. Longitudinal studies are needed to con®rm the observed cross-sectional associations.