Objective: To study the association between baseline vitamin D status, bone mineral density (BMD),, and the development of radiographic osteoarthritis (ROA) of the knee in a large population-based cohort of men and women. Methods: A sample of 1248 subjects (728 women and 520 men) was drawn from the Rotterdam Study, a prospective population-based cohort study of the elderly. At baseline, vitamin D dietary intake was determined, and BMD and 25-hydroxy vitamin D (25(OH)D) serum levels were measured. After a mean follow-up time of 6.5 years incidence and progression of knee ROA of was assessed. Results: The mean vitamin D intake in our study population was 64 IU/d and the mean 25(OH)D level 66 nmol/L. Vitamin D levels were associated with baseline BMD, particularly in subjects with baseline knee ROA. Progressive ROA occurred in 5.1% of the participants in the highest tertile of vitamin D intake against 12.6% in the lowest tertile, resulting in an adjusted odds ratio of 7.7 (95% CI: 1.3-43.5). Both intake and levels of 25(OH)D were not significantly related to incident ROA. However, we found a significant interaction between vitamin D intake and BMD in the association with incident knee ROA (P = 0.03): in subjects with low lumbar spine BMD at baseline we observe an increasing incidence of knee ROA with decreasing vitamin D intake and serum levels. Conclusions: Low dietary vitamin D intake increases the risk of progression of knee ROA. Particularly in subjects with low baseline BMD, vitamin D status seems to influence the incidence and progression of knee ROA. Thus, improving the vitamin D status in the elderly could protect against the development and worsening of knee OA, especially in those with low BMD.