To determine the influence of exercise training on smoking after acute myocardial infarction (AMI), smoking rates in 42 pre-AMI smokers assigned to exercise training were compared with 26 pre-AMI smokers assigned to no training. Exercise training occurred 3-26 weeks after AMI. The increase in functional capacity in 3-26 weeks was significantly greater in training than in no-training patients: 1.8 vs. 1.2 METs respectively (p less than 0.05). Adherence to exercise training was higher in non-smokers and former smokers than in those who continued to smoke: 89% and 88% vs. 80% respectively (NS). The prevalence of smoking 6 months post-AMI was lower in training than in no-training patients: 31% vs. 39% respectively (NS). Plasma thiocyanates collected on a random sample of 42 patients suggested that 19% of patients who are smoking after MI fail to report doing so. Self-reported cigarette consumption at 28 weeks was half as great in training as in no-training patients: 11 +/- 7 vs. 22 +/- 16 cigarettes per day (p less than 0.03). Firm advice to stop smoking followed by medically supervised exercise training with frequent followup reduces self-reported cigarette consumption in patients after AMI.