Nitroglycerin has been shown to cause a downward shift in the left ventricular (LV) pressure-volume relation in patients. To test the hypothesis that this shift is mediated by an alteration in pericardial pressure, 13 patients undergoing diagnostic cardiac catheterization were studied. LV and right ventricular (RV) pressure (micromanometers) and LV diameter (2-dimensional echocardiography) were measured simultaneously before and after sublingual administration of 0.3 to 0.6 mg of nitroglycerin. In the 11 patients with hemodynamic effects from nitroglycerin, mean LV end-diastolic pressure decreased from 12.7 +/- 5 mm Hg (mean +/- standard deviation) to 7.3 +/- 3 mm Hg (p less than 0.002) and mean RV end-diastolic pressure declined from 7.7 +/- 3 mm Hg to 5.0 +/- 1 mm Hg (p less than 0.001). However, nitroglycerin caused only a slight (6%) reduction in LV minor axis diameter, from 52 +/- 8 mm to 49 +/- 9 mm (p less than 0.05). Diastolic pressure-diameter plots constructed from early and late diastolic measurements demonstrated a downward shift in the relation. However, when RV end-diastolic pressure was used as an estimate of pericardial pressure (a procedure validated by studies in our laboratory), the transmural pressure-diameter points before and after administration of nitroglycerin defined a single curve. These observations are in keeping with the conclusions that nitroglycerin did not alter the elastic properties of the myocardium and that the decrease in LV end-diastolic pressure induced by nitroglycerin was primarily attributable to a reduction in external constraint.