To determine whether pericardial liquid pressure accurately measures pericardial constraint, we developed a technique in which a catheter was positioned perpendicular to the epicardial surface. This device, which occupies little or no pericardial space, couples the thin film of liquid to a transducer. In six open-chest dogs, we also measured left ventricular (LV) end-diastolic pressure (LVEDP) and anteroposterior and septum-to-free wall diameters. LVEDP was raised incrementally to ~25 mmHg by saline infusion. With the use of the product of the two diameters as an index of area (A_LV), LVEDP-A_LV relationships were obtained with the pericardium closed and again after the pericardium had been widely opened to obtain the isovolumic difference in LVEDP (LVEDP). In all dogs, the technique yielded values of pericardial pressure equal to LVEDP as well as equal to that measured using a previously placed balloon transducer in the same location and at the same A_LV. We conclude that, when the pressure of the pericardial liquid is appropriately measured, it (in addition to the balloon-measured contact stress) defines the diastolic constraining effect of the pericardium. Furthermore, we suggest that earlier measurements of pericardial "liquid pressure" were low, due to an artifact of measurement.