To compare the measurement of left ventricular ejection fraction obtained by two-dimensional echocardiography and by radionuclide ventriculography in patients following acute myocardial infarction, 49 consecutive patients with acute myocardial infarction underwent echocardiography and radionuclide ventriculography on the same day, pre-discharge. Left ventricular ejection fraction was assessed by two blinded observers for each method and reproducibility was also assessed for each technique. The limits of agreement for the differences in ejection fraction (%) between the two methods was--11.4, 12.2; the mean difference 0.4 was not significantly different from zero. The limits of agreement for the intra- and inter-observer differences in ejection fraction by radionuclide ventriculography were--9.4, 7.6 and -8.6, 11.0, respectively; the mean differences--0.9 and 1.2 were not significantly different from zero. The limits of agreement for the intra- and inter-observer differences by echocardiography were--5.8, 6.6 and--8.9, 9.5 respectively; the mean differences 0.4 and 0.3 were not significantly different from zero. Thus, two-dimensional echocardiography compares well with radionuclide ventriculography for the assessment of ejection fraction without the disadvantage of radiation.