This study examines the reproducibility of individual radionuclide attenuation factors used in the calculation of cardiac output and left ventricular volume by the nongeometric radionuclide method. Twenty male patients were studied at rest with thermodilution measurements of cardiac output on two separate days. Simultaneous equilibrium radionuclide angiograms were performed and left ventricular stroke volume and cardiac output were determined by the nongeometric method. Individual patient attenuation factors were calculated as the ratio of thermodilution and radionuclide cardiac output measurements at each study. There was a close linear relationship between radionuclide and thermodilution measurements of cardiac output in each study (r = 0.88 study 1, r = 0.97 study 2). A similar relationship was found for measurements of left ventricular stroke volume (r = 0.86, study 1, r = 0.97 study 2). Individual radionuclide attenuation factors ranged from 2.49 to 3.46 in study 1 and from 2.77 to 3.29 in study 2. The individual attenuation factors were reproducible to within 10% in 13 patients and to within 15% in 19 patients. When cardiac output was calculated from the radionuclide data of study 2, by means of individual attenuation factors previously determined in study 1, there was a good correlation with the simultaneous thermodilution measurements of cardiac output (r = 0.92, SEE = 0.38 L/min). Individual radionuclide attenuation factors show little variation in serial studies. Thus the nongeometric radionuclide technique can be used to make accurate serial measurements of cardiac output and left ventricular volume.