Over 30 months 9292 consecutive patients admitted to nine coronary care units with suspected myocardial infarction were considered for admission to a randomised double blind study comparing the effect on mortality of nifedipine 10 mg four times a day with that of placebo. Among the 4801 patients excluded from the study the overall one month fatality rate was 18.2% and the one month fatality rate in those with definite myocardial infarction 26.8%. A total of 4491 patients fulfilled the entry criteria and were randomly allocated to nifedipine or placebo immediately after assessment in the coronary care unit. Roughly 64% of patients in both treatment groups sustained an acute myocardial infarction. The overall one month fatality rates were 6.3% in the placebo treated group and 6.7% in the nifedipine treated group. Most of the deaths occurred in patients with an in hospital diagnosis of myocardial infarction, and their one month fatality rates were 9.3% for the placebo group and 10.2% for the nifedipine group. These differences were not statistically significant. Subgroup analysis also did not suggest any particular group of patients with suspected acute myocardial infarction who might benefit from early nifedipine treatment in the dose studied.