A register of acute coronary events in Auckland, New Zealand collected data on both cases of definite myocardial infarction not dying suddenly, and cases of sudden cardiac death. This paper compares these two syndromes with respect to several clinical and pathologic variables. The following variables discriminated between these two syndromes significantly: season of the year, painful prodromal symptoms, use of digoxin and/or furosemide, age, race, alcohol consumption over the previous one year. Possible biases and sources of confounding are discussed. Heavy alcohol consumers have a higher proportion of coronary events as sudden deaths. This proportionate increase is probably due, in part at least, to an absolute increase in the frequency of sudden death. The data are supportive of the idea that persons with a high alcohol consumption who also suffer myocardial infarction are more likely to die suddenly with the infarction. The rather conflicting data from many other studies are discussed. A model is proposed, based on reported data, that could explain the present findings and those previously reported by others.