We hypothesize that 1) both high and low dose B vitamin therapy will lower homocysteine in patients anticoagulated with warfarin and 2) high dose B vitamin therapy will lower thrombomodulin in patients anticoagulated with warfarin. To accomplish these specific aims, we proposed a double- blind randomized controlled trial of 6 months duration comparing 2 different multivitamin regimens differing only in the doses of the following 3 vitamins. The multivitamin for the control group has no folic acid, B2, B6, or B 12. The multivitamin for the intervention group has 5 ing folic acid, 1.7 mg B2, 100 mg B6, and I mg B12. Participants age 50 and older, 150 in each group, are being recruited from the Anticoagulation Clinics at the Baltimore VAMC and neighboring centers. Clinical and risk factor information is being obtained through a face-to-face interview. Blood is being drawn for analyses of homocysteine and vitamin levels (University of Colorado Health Science Center) and thrombomodulin (Baltimore VAMC Hemostasis and Thrombosis Laboratory). Data analysis will be conducted according to the "intention to treat" principle using ordinary least squares regression models.