The aim of this study was to analyse the influence of patient characteristics on delay between onset of symptoms and hospital admission (patient delay) in acute myocardial infarction. A group of 6676 consecutive patients with AMI, admitted alive to 27 Danish hospitals from 1990 to 1992, were studied. Due to missing information on delay or in hospital acute myocardial infarction 698 patients were excluded. Mean patient delay was 9.1 hours, median delay 3.25 hours (5 to 95 percentiles: 0.67-40 hours). In multivariate logistic regression analysis patient delay was independently associated with male gender, increased age, diabetes mellitus, left ventricular systolic function (wall motion index), onset from midnight to 6 a.m., onset on a weekday, history of angina pectoris, chest pain as initial symptom, ventricular fibrillation or-tachycardia, Killip class > or = 3, presence of ST-elevation and ST-depressions. In conclusion, patient delay continues to be disappointingly long. This also applies to patients with a high risk of acute myocardial infarction (notably history of diabetes mellitus and angina pectoris).