Few data exist on the association between vitamin K status and bone mineral density (BMD) in men and women of varying ages. We examined cross-sectional associations between biochemical measures of vitamin K status and BMD at the hip and spine in 741 men and 863 women (mean age, 59 yr; range, 32-86 yr) who participated in the Framingham Heart Study (1996-2000). Vitamin K status was assessed by plasma phylloquinone and percentage undercarboxylated osteocalcin (%ucOC). Among the men, low plasma phylloquinone concentrations adjusted for triglycerides and elevated serum %ucOC levels were associated with low BMD at the femoral neck (P = 0.03 and 0.009, respectively). Among postmenopausal women not using estrogen replacements, low plasma phylloquinone concentrations were associated with low spine BMD (P = 0.007), with a nonsignificant trend of an elevated serum %ucOC with low spine BMD (P = 0.08). In contrast, there were no significant associations between biochemical measures of vitamin K and BMD in either premenopausal women or postmenopausal women using estrogen replacements. Clinical trials are required to isolate any putative effects of vitamin K on rates of bone loss. The target population in these trials, particularly in regard to estrogen use, may be critical, as suggested by the findings of this study.