Systematic exercise training effectively improves exercise capacity in patients with coronary artery disease (CAD), but the magnitude of improvements is highly heterogeneous. We investigated whether this heterogeneity in exercise capacity gains is influenced by the insertion/deletion (I/D) polymorphism of the angiotensin-converting enzyme (ACE) gene. Patients with CAD (n = 169) were randomly assigned to 12 weeks of exercise training or standard care, and 142 patients completed the study. The ACE polymorphism was determined for 128 patients (82% males, 67 ± 9 years). Peak oxygen uptake was measured before and after the 12-week intervention. The ACE I/D polymorphism frequency was n = 48 for D/D homozygotes, n = 61 for I/D heterozygotes and n = 19 for I/I homozygotes. Baseline peak oxygen uptake was 23.3 ± 5.0 ml/kg/min in D/D homozygotes, 22.1 ± 5.3 ml/kg/min in I/D heterozygotes and 23.1 ± 6.0 ml/kg/min in I/I homozygotes, with no statistical differences between genotype groups (P = 0.50). The ACE I/D polymorphism frequency in the exercise group was n = 26 for D/D, n = 21 for I/D and n = 12 for I/I. After exercise training, peak oxygen uptake was increased (P < 0.001) in D/D homozygotes by 2.6 ± 1.7 ml/kg/min, in I/D heterozygotes by 2.7 ± 1.9 ml/kg/min, and in I/I homozygotes by 2.1 ± 1.3 ml/kg/min. However, the improvements were similar between genotype groups (time × genotype, P = 0.55). In conclusion, the ACE I/D polymorphism does not affect baseline exercise capacity or exercise capacity gains in response to 12 weeks of high-intensity exercise training in patients with stable CAD. Clinical trial registration: www.clinicaltrials.gov (NCT04268992).