Factors associated with delay to hospital arrival after the onset of symptoms suggestive of acute myocardial infarction (AMI) were examined in the late 1960s and 1970s, but recent data concerning these characteristics are limited. The purpose of the present study was to examine overall and temporal distributions of the extent of patients' delay from the time of onset of AMI symptoms to hospital arrival and factors associated with delay in seeking medical care from a multihospital, population-based perspective. Review of medical records was undertaken of patients hospitalized with a discharge diagnosis of AMI in 16 teaching and community hospitals in Worcester, Mass. in 1986, 1988, and 1990. The study sample comprised 1279 patients hospitalized with validated AMI in whom data concerning extent of patient delay from onset of symptoms suggestive of AMI to hospital arrival were available. The average delay between onset of symptoms suggestive of AMI and arrival at local emergency departments did not change significantly with time (average of 4.1 hours in 1986, 4.0 hours in 1988, and 4.6 hours in 1990). The median delay was 2.0 hours during each of these years. Fifty percent of patients with AMI went to area-wide emergency departments within 2 hours of the onset of acute symptoms, 22% between 2 and 4 hours, and 28% delayed > 4 hours. Results of a multivariable regression analysis showed that older age, history of diabetes, type of medical insurance coverage and previous AMI were significantly associated with delays in hospital arrival of > 2 hours.(ABSTRACT TRUNCATED AT 250 WORDS)