Observational and clinical trial experience suggest the possibility that magnesium may have both a physiologic and pharmacologic role in patients with essential hypertension. Unfortunately, the evidence from epidemiologic studies is inconsistent and many of the clinical trials are small and methodologically imperfect. In theory, magnesium supplementation could be used as a nonpharmacologic hypotensive agent in the individual patient with established hypertension or as a population-based treatment strategy aimed at achieving a slight downward shift in the entire distribution of blood pressure. Our current level of uncertainty is likely to persist until results from large, rigorously controlled observational and interventional studies become available.