Objective: Vitamin D deficiency and polypharmacy are common in the elderly. However, knowledge on the associations between the use of specific medicines and serum 25-hydroxyvitamin D (25(OH)D) is limited. The aim of this study was to (better) define the associations between the use of specific medicines and serum 25(OH)D. Methods: Two different cohorts (1995/1996 and 2002/2003) from the Longitudinal Aging Study Amsterdam (LASA) were used for cross-sectional analyses. LASA is based on an age and sex-stratified random sample of the Dutch older population. Study participants were aged 65-88 years in the first cohort (nZ1301) and 55-65 years in the second cohort (nZ736). Serum 25(OH)D of users of several groups of medicines were compared with levels of non-users using multiple linear regression analysis. Results: Of all participants, 75.4% (first cohort) and 61.1% (second cohort) were using at least one medicine. In both cohorts, the number of medicines was associated with lower serum 25(OH)D. In the first cohort, after adjustment for confounding, users of any kind of medicine, loop diuretics and inhaled corticosteroids (only men) had respectively 4.4 nmol/l (P<0.01), 4.7 nmol/l (PZ0.04) and 7.3 nmol/l (PZ0.02) lower serum 25(OH)D than non-users. In the second cohort, the use of oral antidiabetics, calcium-channel blockers and angiotensin-converting enzyme inhibitors was associated with respectively 7.4 nmol/l (PZ0.04), 7.7 nmol/l (PZ0.01) and 7.6 nmol/l (P<0.01) lower serum 25(OH)D. Conclusions: These data show that users of several medicines have lower serum 25(OH)D than non-users. Vitamin D supplementation may be considered in patients with chronic use of medicines. © 2012 European Society of Endocrinology.