Smoking is an important risk factor for atherosclerotic heart disease, but several studies have shown smoking to be associated with a favourable prognosis in patients who have suffered an acute myocardial infarction (AMI). We studied a large group of consecutive patients admitted alive to hospital with an infarction in order to further study the prognostic importance of smoking status at the time of myocardial infarction. The study cohort comprised 6676 patients with an enzyme-confirmed myocardial infarction admitted to 27 Danish hospitals over a 26-month period between 1990 and 1992. Smoking status was determined at the time of hospitalisation and complete follow-up was obtained in October 1996. Smokers were on average 10 years younger, had fewer concomitant cardiac risk factors, and were more likely to be male and to receive thrombolytic therapy more frequently than non-smokers. In univariate analysis, smoking was associated with reduced 30-day and long-term mortality (risk ratio at 30 days 0.55, P < 0.001, risk ratio long-term 0.59, P < 0.001). When age only was included in a multivariate analysis, smoking was no longer of importance in short- or long-term mortality (risk ratio 0.92, P = 0.4 at 30 days and long-term risk ratio 0.98, P = 0.7). Inclusion of further variables did not change this picture. In conclusion, smoking contributes to the occurrence of AMI at a younger age. The more favourable prognosis in smokers at the time of AMI is a result of more favourable baseline characteristics, especially their lower age.