Gated blood pool angiography was used to study noninvasively the hemodynamic effects of nitroglycerin. In 26 patients with a spectrum of resting left ventricular function abnormalities, serial 1 minute determinations of count-derived ejection fraction, end-diastolic volume, end-systolic volume and cardiac output (radionuclide stroke volume × heart rate), as well as heart rate and blood pressure, were compared for 6 minutes before and 14 minutes after the administration of 0.6 mg of sublingual nitroglycerin. For the entire group, mean ejection fraction increased 15.6 ± 2.2 percent (mean ± standard error of the mean), heart rate increased 6.1 ± 0.6 percent, systolic blood pressure decreased 11.2 ± 1.3 percent, end-diastolic volume decreased 10.7 ± 2.1 percent, end-systolic volume decreased 18.5 ± 2.6 percent (all P <0.001), and radionuclide cardiac output increased 5.3 ± 3.8 percent (P = not significant). There was a limited correlation between changes in ejection fraction and changes in radionuclide cardiac output (r = 0.65) and an inverse correlation between basal ejection fraction and change in radionuclide cardiac output (r = -0.64). Three patients had a significant decrease in radionuclide cardiac output with no change in ejection fraction. Changes in systolic blood pressure also correlated poorly with both ejection fraction (r = 0.3) and radionuclide cardiac output (r = 0.126). It is concluded that radionuclide angiography may be useful in documenting noninvasively the efficacy or adverse effects of vasodilator agents. Changes in cardiac output were not predicted well by other variables and must be calculated directly. © 1979.