BACKGROUND--The exact mechanism that explains the phenomenon of cold intolerance in patients with angina remains controversial. Although the response to the effects of a cold environment has been examined in these patients, their response to cold air inhalation has produced conflicting results. In addition, the possible role of vasoactive peptides in the pathophysiology has not been explored. OBJECTIVES--The aims of this study were to examine the response of patients with stable angina to the effects of cold air inhalation during exercise testing, and to investigate the possible role played by the vasoconstrictor peptides endothelin-1 (ET-1) and angiotensin-II (AT-II) in the pathophysiology. METHODS--In a randomised order, 12 men with stable angina, whose medication had been stopped, underwent two separate symptom limited treadmill exercise tests. At one visit the patients exercised while breathing room air and at the other visit they exercised while breathing cold air from a specially adapted freezer. Serial peripheral venous blood samples were taken for ET-1 and AT-II estimations during each visit. RESULTS--Cold air inhalation resulted in a significant reduction in the mean time to angina (232.7 (20.4) s v 274.1 (26.9) s, P = 0.04) and the mean total exercise time (299.5 (27.0) s v 350.3 (23.9) s, P = 0.008), but no significant change in the time to 1 mm ST depression (223.3 (29.0) s v 241.3 (29.2) s, P = 0.25). There was no significant difference between the rate-pressure products at the onset of angina (P = 0.13) and the time to 1 mm ST depression (P = 0.85), but at peak exercise the rate-pressure product was significantly lower in patients breathing cold air as opposed to room air (P = 0.049). There was an equivalent significant decrease in ET-1 concentrations at peak exercise compared with that at rest at both visits (room air 5.0 (0.7) pmol/l v 4.3 (0.7) pmol/l, P = 0.03; cold air 4.4 (0.6) pmol/l v 3.8 (0.5) pmol/l, P = 0.02). There was a significant increase in AT-II concentrations 10 min after peak exercise in patients breathing room air (39.2 (6.1) pmol/l v 32.1 (4.8) pmol/l, P = 0.01) which was not repeated during cold air inhalation (36.6 (3.4) pmol/l v 28.3 (3.4) pmol/l, P = 0.07). CONCLUSIONS--Cold air inhalation in patients with stable angina results in an earlier onset of angina and a reduction in exercise capacity. Both peripheral and central reflex mechanisms appear to contribute to the phenomenon of cold intolerance. Peripheral ET-1 and AT-II do not appear to play a significant role in the pathophysiology.