Past studies of the incidence of acute myocardial infarction (AMI) have found information for patients managed at home difficult and expensive to obtain. This study attempted to find whether the efficiency of gathering this information could be improved, or if this group of patients could be neglected entirely. A random sample of 270 doctors in Perth were asked about their management of cases of suspected AMI at home. Of these doctors, 78% said that they never managed AMI at home; excluding such doctors before seeking voluntary notification of new cases managed at home could make population surveys of AMI more efficient. For patients under 65 it was estimated that about 1% or less of non-fatal cases meeting World Health Organization criteria for definite AMI were managed at home. Thus neglecting this group would probably cause little bias in computing incidence rates for AMI in Perth.