Despite improvements in the outcome of patients with acute myocardial infarction during the last three decades, room for improvement exists in the elderly patients and in patients who are not candidates for throm-bolysis. Statistical analysis of randomized trials of magnesium in myocardial infarction reveals a gradient of response. When higher risk patients were enrolled, a greater benefit of magnesium was observed; progressively smaller benefits of magnesium occur as control group mortality approached 7%, at which point no benefit was detected. Although the ISIS-4 study enrolled more than 58,000 patients, no reduction in mortality was seen, probably as a result of low control group mortality and relatively late administration of magnesium. Because the potential benefits of magnesium in myocardial infarction remains an open question, additional trials are needed before this inexpensive and early-administered therapy is prematurely cast aside.