We re-examined, in the context of modern practice, plasma insulin and stress hormone concentrations in patients admitted to hospital with acute coronary syndromes. Venous blood sampling was carried out prior to anti-thrombotic therapy in 148 patients with myocardial infarction (MI); 76 patients with unstable angina (UA) pectoris were also studied, together with 27 patients with non-cardiac chest pain (NCP). There were significant progressive increases in the concentrations of catecholamines, cortisol, glucose and insulin from NCP to UA to MI patients. Hyperglycaemia (glucose >8 mmol/l) was present in over 50% of MI patients. The plasma cortisol and insulin levels were both significantly positively correlated with the glucose concentration on admission. Only the cortisol concentration was correlated with peak cardiac enzyme levels. The glucose and insulin concentrations on admission in 141 MI and UA patients were related to insulin resistance, as judged from subsequent insulin and glucose concentrations measured while fasting and during a glucose tolerance test. The product of admission insulinxglucose (divided by 25; the admission insulin-resistance index, or AIRI) was significantly correlated with indices of insulin resistance, and was significantly higher (approximately double) in the MI group (7. 81+/-0.76) and the UA group (6.88+/-1.19) than in the control NCP group (3.59+/-0.06; Kuskul-Wallis: P=0.0001), implying that the insulin levels in the first two groups were approximately twice as high as is appropriate for the glucose levels. The ethnic origin of 20% of the patients was the Indian subcontinent; admission insulin and glucose levels in this subgroup were higher than in the non-Asians across all the groups with chest pain. Cortisol was the only stress hormone that was raised in proportion to the size of the infarct, and is a likely partial cause of the elevation in blood glucose. The high insulin levels were related to the prevalence of insulin resistance, and this was particularly important in the Asian subgroup presenting with MI and UA. Thus it appears feasible to identify acute coronary syndrome patients who are insulin-resistant at a time (on admission) when alternative early therapeutic strategies can be instituted.