1 We have performed a double-blind, randomized, 7-d cross-over study of the thromboxane synthetase inhibitor dazoxiben (UK 37248) in 20 patients with stable coronary heart disease. 2 All patients had a history of exertional angina of greater than two years duration and no patient had suffered a myocardial infarction in the preceding twelve months. 3 All patients had a positive exercise stress test for myocardial ischaemia and 15 had undergone coronary angiography. All these patients had a 50% narrowing in at least one vessel. 4 All patients were on conventional anti-anginal medication and the doses of their various therapies remained unchanged in the three months prior to and during the study period. 5 Therapy with dazoxiben 200 mg four times daily produced no alteration in the subjective or objective features of angina in these patients. There was no alteration in angina attack rate, glyceryl trinitrate consumption or duration of treadmill exercise. 6 Dazoxiben produced a highly significant reduction in both the resting and the post-exercise levels of serum thromboxane B2 levels, although there was no significant difference between the pre-exercise and post-exercise values. 7 Dazoxiben is an effective inhibitor of the synthesis of thromboxane but it has no effect on the subjective or objective features of stable coronary disease. This would suggest that the production of thromboxane and the development of circulating platelet aggregates play no part in the mechanism of angina in patients with stable coronary heart disease.