Left ventricular end-systolic pressure-volume relation has been proved experimentally to be an useful index of left ventricular contractility relatively independent of preload or afterload. But less clinical application has been reported because of its invasive nature, and we evaluated this relationship non-invasively using gated radionuclide angiocardiography as volume determination and cuff sphyngomanometer in the arm as pressure measurement. Gated equilibrium blood pool scintigrams were obtained at rest and during intravenous infusion of angiotensin or nitrate. Ventricular volumes were derived from ventricular activity and peripheral blood volume and activity. The peak systolic pressure (PSP) by cuff method to end-systolic volume index (ESVI) relations showed good linearity (r greater than .930 in 84% of consecutive 50 cases) and were gentler in the groups with more impaired left ventricular function. Emax was related exponentially to ejection fraction (EF) and hyperbolically to end-diastolic volume index. The dead volume (VoI) was unfixed and fell into positive or negative value, and was not related to EF under control condition. PSP/ESVI in each loading condition was less variable with the alteration of blood pressure than EF. The linear relation was found between PSP/ESVI under control condition and Emax (PSP/ESVI = 0.651 . Emax + 0.958, r = 0.841, p less than .001). Thus in measuring ventricular volume, gated radionuclide angiocardiography is a non-invasive method less affected by the geometry of the left ventricle. Non-invasive determination of end-systolic pressure-volume relation using the volume by radionuclide and the blood pressure by cuff method is clinically useful in the assessment of left ventricular contractility.