The prevalence of osteoarthritis (OA) is greater in women then men. Weight, a factor strongly associated with osteoarthritis, is significantly increased over the menopausal transition. Despite the high prevalence of osteoarthritis, a disabling disease with limited treatment options, there is a paucity of studies in women. The longitudinal phase of the Melbourne Women's Midlife Health Project, is a population-based prospective study of 438 Australian born women who have been followed annually over 11 years. 257 (59%) of these women remained in longitudinal assessment at 11th year of follow-up and 224 of these women agreed to undergo X-rays of their knees and hands. In this study, X-rays were scored for evidence of osteoarthritis using a validated scale, by two investigators who were blinded to questionnaire results. Information on hormone therapy use, physical activity, smoking, BMI and age were obtained by both self-administered and face-to-face questionnaires. Results showed that one hundred and twenty-eight women (56%) had evidence of radiological OA. Forty-nine (21.6%) had evidence of radiological knee OA. One hundred and one (44.5%) had evidence of radiological hand OA. Compared to baseline, at 11th year of follow-up participants had put on weight (4 kg range -14 to 25 kg) and a larger proportion (25% vs. 40%) reported exercising frequently (P = 0.005). Physical activity and BMI were associated with higher prevalence of osteoarthritis in the final year of follow-up. Smoking was associated with a lower prevalence of radiological knee OA. The prevalence was 61% among never smokers compared to only 39% among those who smoked (P < 0.05). Total OA was associated with never having used hormone therapy (odds ratio 2.7; CI 1.1-6.9). There was a trend for increasing level of physical activity at ages 20-29 to be associated with an increased risk of knee OA (P value 0.03 for trend). In conclusion increasing age, BMI and history of more frequent physical activity in younger years were risk factors for radiological knee OA. In contrast, smoking appeared to be associated with less knee OA. Never having used hormone therapy was a risk factor for radiological hand and knee OA. Further work will be needed to determine whether modification of these factors can prevent the development of OA.