Total numbers of daily hospital admissions for cardiac emergencies were obtained from 3 hospitals within the Sudbury Basin in Canada for November through March for each of 4 consecutive years (1983-1986). Major diagnostic categories were also differentiated. No statistically significant correlations were found between the amount of snowfall during the day of or the days before or after admissions for the major categories of cardiac emergency. Specific analyses, conducted to reveal possible recondite associations between extreme or cumulative snowfalls and the most extreme days of cardiac emergencies, indicated a chance association. We conclude that the occurrence of heart attacks is independent of snowfall but that, when they occur, they are attributed to shoveling if there has been a recent major snow storm.