SUMMARY Frame-by-frame analysis of angiograms in 16 patients revealed that hemodynamic interventions are capable of producing substantial shifts in the diastolic pressure-volume curve. Angiotensin raises blood pressure and shifts the entire pressurevolume curve up, and nitroprusside lowers blood pressure and shifts the curve down. Indirect measurements of pleural pressure in seven patients (via esophageal pressure) showed that pleural pressure changes were too small to account for these shifts. Analyzing our results in terms of a theoretical pressure-volume equation previously validated in dog studies did not show the observed shifts to be the product of acute changes in the elasticity of the myocardium itself. This same analysis suggested that indirect changes in the external mechanical constraints acting on the left ventricle such as the right ventricular pressure, the pericardium, and perhaps viscoelastic effects related to changes in filling rate account for the pressure-volume curve shifts with intervention. The fact that one cannot in general relate a specific volume to a given pressure in the face of hemodynamic interventions calls into question the use of end-diastolic pressure interchangeably with end-iastolic fiber length when interpreting systolic events in terms of the Frank-Starling mechanism.DURING DIASTOLE the left ventricle fills passively, with its pressure and volume related by a curve which reflects both ventricular geometry and the elasticity or stiffness of cardiac muscle. One generally describes the passive elasticity of cardiac muscle by means of an exponential curve relating stress (force/area) to strain (percentage distension), with the muscle becoming stiffer as it is stretched.' The pressure-volume relationship has been assumed to be sufficiently constant over the short term to use left ventricular end-diastolic pressure interchangeably with enddiastolic circumferential fiber length in analyses of ventricular function based on the Frank-Starling mechanism. Chronic cardiac disease may cause the ventricle to dilate or hypertrophy, and these geometric changes distort its diastolic pressure-volume relationship.2 In addition, disease might change the passive elasticity of the myocardium itself, altering the pressure-volume relationship.3-6 In an acute situation, however, it is assumed that hemodynamic interventions simply move the ventricle up and down the same pressure-volume curve. This assumption is based on the fact that no agents are known to affect passive elasticity of normally oxygenated papillary muscle,' and the fact that the ventricle cannot acutely become hypertrophied.We explored this assumption of constancy of the pressure-volume relationship by measuring diastolic pressurevolume curves based on frame-by-frame analysis of angiograms in 16 patients. Acute interventions with nitroprusside and angiotensin usually noticeably shifted the diastolic pressure-volume curve. This result brings into question the assumption that the diastolic pressure-volume relationship remains constant ...