Cross-sectional studies suggest an association between BMI and pain. This prospective study investigated the associations of measured BMI and waist circumference with prevalent and incident pain in older adults. The study included participants of the Longitudinal Aging Study Amsterdam, aged 55-85 years at baseline (1992-1993). Pain was assessed using a subscale of the Nottingham Health Profile at baseline (N = 2,000), after 3 years (N = 1,478) and 6 years (N = 1,271) of follow-up. The overall prevalence of pain was 32.7% at baseline and increased significantly with higher quartiles of BMI or waist circumference. After adjustment for age, education, depression, smoking, physical activity, and chronic diseases, multiple logistic regression analyses showed odds ratios (ORs (95% confidence interval)) for prevalent pain of 2.16 (1.32-3.54) in men and 1.93 (1.26-2.95) in women comparing the highest with the lowest quartile of BMI. Of the participants without pain at baseline, those in the highest quartile of BMI had a twofold increased odds for incident pain after 3 years of follow-up. After 6 years of follow-up, ORs for incident pain were 2.34 (1.17-4.72) in men and 2.78 (1.36-5.70) in women. Additional adjustment for weight change did not change these associations. Similar results were found for the associations between waist circumference and pain. Exploring the reversed causal relation, analyses showed no significant associations between prevalent pain and weight gain. In conclusion, the prevalence of pain is higher among obese older men and women compared to their normal-weight peers. Furthermore, obese older adults are at increased odds to develop pain.