It has recently been suggested that intravenous infusion of magnesium may reduce mortality and the incidence of serious arrhythmias in patients with ischaemic heart disease and acute myocardial infarction. In the present double-blind, placebo-controlled study, 298 patients with suspected acute myocardial infarction were randomized to receive either intravenous magnesium chloride (80 mmol.24 h-1) or placebo. Infusions were started immediately after admission to the coronary care unit. One hundred and fifty patients received magnesium and 148 the placebo. Ischaemic heart disease was diagnosed in 244 patients. Acute myocardial infarction was observed among 83 patients in the magnesium group and 79 in the placebo group. Both treatment groups were comparable regarding sex, age, clinical status, previous cardiac disease and medication. Serum magnesium was significantly raised during magnesium infusion compared to placebo (P less than 0.01). Fatal events were only observed among patients with myocardial infarction, but neither the in-hospital mortality (magnesium: 12.1%; placebo 10.1%) nor the mortality after a follow-up period of 245 days (median observation time) was affected by magnesium substitution. Magnesium infusion was accompanied by a significantly increased incidence of atrioventricular conduction disturbances. The results suggest that patients suffering from acute ischaemic heart syndromes do not benefit from intravenous magnesium supplementation.