INTRODUCTION: Observational data has suggested a link between vitamin D deficiency and coronary heart disease (CHD). However, randomized controlled trials (RCTs) have failed to show benefit. The objective of this study is to analyze the RCTs investigating vitamin D supplementation and the risk of CHD and stroke. METHODS: All RCTs that compared vitamin D supplementation to placebo and evaluated nonfatal myocardial infarction (MI), cardiac mortality, stroke and CHD events (a composite of cardiac mortality, MI, unstable angina and revascularization) were included. Rate ratios (RR) were calculated for each endpoint and to test for heterogeneity of treatment effect (HTE) the Chi(2) and I(2) tests were used for younger vs. older participants, shorter vs. longer trial duration, vitamin D supplements with vs. without calcium, and daily vs. monthly dosages of vitamin D. A meta-regression was performed with baseline vitamin D concentration as the covariate. RESULTS: 22 RCTs were identified (n = 83,200). Vitamin D supplementation had no effect on nonfatal MI (RR 0.98, 95% confidence interval (CI) 0.89-1.08), cardiac death (RR 0.94, CI 0.84-1.06), CHD events (RR 1.00, CI 0.91-1.10), or stroke (RR, 0.97, CI 0.9-1.03). When performing the meta-regression with baseline circulating 25-hydroxyvitamin D (25(OH)D) concentrations as the covariate, vitamin D supplementation's treatment effect on CVD outcomes was not associated with baseline 25(OH)D. CONCLUSION: Vitamin D did not reduce CHD and stroke. A linear relationship does not exist between baseline 25(OH)D and vitamin D supplementation's effect on CVD. Vitamin D levels should be checked and repleted in those with an absolute indication.