Large-scale trials of angiotensin converting enzyme (ACE) inhibitors after acute myocardial infarction (AMI) suggest that the benefits are greatest in patients with left ventricular (LV) dysfunction. However, early evaluation of LV function in all patients after AMI by current methods can be difficult due to a lack of resources and skilled personnel. Thus a clinical algorithm that could be used at the bedside to reliably identify patients with a left ventricular ejection fraction (LVEF) <or = 40% would be helpful as an occasional alternative to echocardiography. We have devised such an algorithm based on the presence of one of: (i) clinical signs of heart failure; (ii) an index Q-wave anterior myocardial infarction; (iii) lack of thrombolytic therapy when there is a history of two or more previous myocardial infarctions and a CK rise > 1000 U/l. We tested this new algorithm prospectively in the coronary care units of two hospitals (one UK and one USA). In the UK centre, the sensitivity and specificity of the algorithm at identifying patients with a LVEF <or = 40% were 82% and 72%, respectively. In the US centre, the sensitivity of the algorithm was 91% and the specificity 78% at identifying patients with LV dysfunction. We have validated a simple clinical algorithm which can be used at the bedside for identifying patients who would benefit from an ACE inhibitor after AMI.