Background: There is still an urgent clinical need to develop non-invasive diagnostic tests for early ischemic heart disease because, once angina occurs, it is too late. Hypocapnia has long been known to cause coronary artery vasoconstriction. Some new cardiology tests are accompanied by the claim that they must have potential diagnostic value if hypocapnia enhances their cardiac effects in healthy subjects. But no previous study has tested whether hypocapnia produces bigger cardiac effects in patients with angina than in healthy subjects. Methods: Severe hypocapnia (a PetCO 2 level of 20 mmHg) lasting >15 min was mechanically induced by facemask, while conscious and unmedicated, in 18 healthy subjects and in 10 patients with angina and angiographically confirmed coronary artery disease, awaiting bypass surgery. Each participant was their own control in normocapnia (where CO 2 was added to the inspirate) and the order of normocapnia and hypocapnia was randomized. Twelve lead electrocardiograms (ECG) were recorded and automated measurements were made on all ECG waveforms averaged over >120 beats. 2D echocardiography was also performed on healthy subjects. Results: In the 18 healthy subjects, we confirm that severe hypocapnia (a mean PetCO 2 of 20 ± 0 mmHg, P < 0.0001) consistently increased the mean T wave amplitude in leads V1-V3, but by only 31% (P < 0.01), 15% (P < 0.001) and 11% (P < 0.05), respectively. Hypocapnia produced no other significant effects (p > 0.05) on their electro-or echocardiogram. All 10 angina patients tolerated the mechanical hyperventilation well, with minimal discomfort. Hypocpania caused a similar increase in V1 (by 39%, P < 0.05 vs. baseline, but P > 0.05 vs. healthy controls) and did not induce angina. Its effects were no greater in patients who did not take β-blockers, or did not take organic nitrates, or had the worst Canadian Cardiovascular Society scores.