Coronary collateral circulation exerts protective effects on myocardial ischaemia due to coronary artery disease and can be promoted by exercise with heparin co-administration. Whether this arteriogenetic effect is accompanied by functional improvement of left ventricle during stress and lessening of angina symptoms remains unknown. To evaluate the anti-ischaemic efficacy of heparin plus exercise in coronary artery disease. In a prospective, single-centre, randomized, double-blind study we recruited 32 'no-option' patients (27 males; mean age 61 ± 8 years) with stable angina, exercise-induced ischaemia and coronary artery disease not suitable for revascularization. All underwent a two-week cycle of exercise (two exercise sessions per day, five days per week) and were randomized (n = 16 per group) to intravenous placebo (0.9% saline) versus unfractionated heparin (5.000 IU intravenously), 10 min prior to exercise. We assessed Canadian Cardiovascular Society angina class, stress electrocardiogram and echo parameters (wall motion score index) and computed tomography angiography for collaterals. After two-week cycle, Canadian Cardiovascular Society class statistically decreased in both groups (heparin plus exercise group: 2.6 ± 0.7 to 1.9 ± 0.7, p < 0.001, exercise group: 2.4 ± 0.7 to 2.1 ± 0.9, p = 0.046). Only the heparin plus exercise group improved time-to-ST segment depression (before 270, 228-327 s vs. after 339, 280-360 s, p = 0.012) and wall motion score index (before 1.38 ± 0.25 vs. after 1.28 ± 0.18, p = 0.005). By multi-slice computed tomography angiography, collaterals improved in 12/15 (80%) in the heparin plus exercise group versus 2/16 (12.5%) in the exercise group (p < 0.001). A two-week, 10-test cycle of heparin plus exercise is better than exercise in improving angina class, myocardial ischaemia and collaterals by computed tomography angiography.