Magnesium deficit may be considered as a cardiovascular risk because of its aetiopathogenic role in the genesis of atherogenous dyslipidaemias and the so-called "idiopathic" mitral valve prolapse. It does not, however, constitute a major antihypertensive factor, though it may sometimes be an accessory co-factor. Plasma magnesium is generally normal in untreated hypertensive patients and normotension is the rule during magnesium deficit. An inverse relationship between magnesium and renin in the plasma of hypertensives has not been confirmed. In practice, plasma magnesium seems to be related to the evolution of the disease. An inverse correlation between blood pressure and erythrocyte total and free magnesium levels has been observed in diverse selected populations but no adjustment has been made in these studies for important covariables. A weak positive association between blood pressure and erythrocyte free magnesium was lost in a multivariate regression analysis. As a rule there is no difference between erythrocyte, leucocyte, and lymphocyte magnesium in hypertensives and controls. More often no relation between urinary magnesium and blood pressure is observed. Daily urine magnesium may be increased with increased excretion of urine adrenaline. Epidemiological data on dietary magnesium, particularly in drinking water, should be carefully scrutinized: these studies do not establish a major role for magnesium as an antihypertensive factor but confirm the importance of magnesium deficit as a nephrocardiovascular risk factor and sometimes gives support for a role of magnesium as an antihypertensive cofactor. The use of magnesium-depleting drugs in hypertensive patients may induce magnesium depletion which must be palliated.(ABSTRACT TRUNCATED AT 250 WORDS)