Universal reduction in sodium intake has long been recommended, largely because of its proven ability to lower blood pressure for some. However, multiple randomized trials have also demonstrated that similar reductions in sodium increase plasma renin activity and aldosterone secretion, insulin resistance, sympathetic nerve activity, serum cholesterol, and triglyceride levels. Thus, the health consequences of reducing sodium cannot be predicted by its impact on any single physiologic characteristic but will reflect the net of conflicting effects. Some 23 observational studies (>360,000 subjects and >26,000 end points) linking sodium intake to cardiovascular outcomes have yielded conflicting results. In subjects with average sodium intakes of less than 4.5 g/day, most have found an inverse association of intake with outcome; in subjects with average intakes greater than 4.5 g/day, most reported direct associations. Finally, in two, a "J-shaped" relation was detected. In addition, three randomized trials have found that heart failure subjects allocated to 1.8 g of sodium have significantly increased morbidity and mortality compared with those at 2.8 g. At the same time, a randomized study in retired Taiwanese men found that allocation to an average intake of 3.8 g improved survival compared with 5.3 g. Taken together, these data provide strong support for a "J-shaped" relation of sodium to cardiovascular outcomes. Sodium intakes above and below the range of 2.5-6.0 g/day are associated with increased cardiovascular risk. This robust body of evidence does not support universal reduction of sodium intake.