Although there is considerable epidemiologic evidence for a relation between plasma homocysteine (HCY) and cardiovascular (CV) disease, the role of HCY as a causal CV risk factor remains controversial, mainly because of the intercorrelation between HCY and other CV risk factors. The goal of the present nested case-control prospective study is to determine the multiadjusted relation between HCY and CV mortality in a large and low CV risk population after a mean follow-up of 14 years. In 1980 and 1981, plasma was saved from 5,000 patients who underwent a systematic health checkup, including clinical and biologic examinations. In 1999, HCY concentration was measured in 110 subjects who died of CV disease (cases) and in 154 randomly matched survivors (control subjects). Statistical analysis was adjusted for CV risk factors. Based on Cox analyses, 3 factors emerged as independent predictors of CV mortality: C-reactive protein, systolic blood pressure, and HCY. The adjusted hazard ratio for CV mortality was 1.22 (95% confidence interval 1.04 to 1.41) per 1 SD (3.9 micromol/L) increment of HCY. Thus, HCY is an independent risk predictor for CV mortality. Because of extensive adjustment procedures, the present study provides additional epidemiologic evidence for a causal relation between HCY and CV disease.