Radionuclide left ventricular (LV) peak filling rate (PFR) was determined in 185 survivors of acute myocardial infarction (AMI) and expressed in units of (1) end-diastolic volume per second (EDV s-1). (2) stroke volume per second (SV s-1), or (3) actual millilitres of blood filled into the left ventricle per second (ml s-1). The purpose of the study was to assess the interrelationship between the three expressions of PFR, and to analyse their significance with regard to signs of congestive heart failure and 1-year survival in patients with AMI. PFR EDV s-1, PFR SV s-1 and PFR ml s-1 had a poor relationship to each other, were all influenced by LV volumes and ejection fraction, and supplied contradictory information with regard to LV filling in patients with heart failure. None of the three expressions of LV peak filling rate had an association to heart failure that was independent of LV volume and ejection fraction. A low PFR EDV s-1 in contrast to a high PFR SV s-1 was associated with a high 1-year cardiac mortality, suggesting that these 'normalized' indices of LV peak filling rate signalled LV size and stroke volume rather than actual LV filling behaviour. No association was present between PFR ml s-1 and 1-year mortality. We conclude that the clinical use of radionuclide LV PFR in patients with AMI may lead to spurious results, unless the influence of LV size and ejection fraction is taken into consideration.