Pulse-transit sonomicrometry was used to measure the base-apex (a), anteroposterior (b), and septal-free wall (c) diameters of the left ventricle and the septal-free wall diameter of the right ventricle (d) in eight excised and three isolated, pump-perfused canine heart preparations, as well as in nine conscious dogs. In the three perfused hearts and in four of the excised hearts, right ventricular free wall regional segment lengths and segment area also were assessed. Biventricular volumes were measured directly with intracavitary balloons in all isolated hearts. When left ventricular balloon volume was held constant, relations between right ventricular free wall dimensions and right ventricular balloon volume were highly linear. With increments in left ventricular volume, however, these relations remained linear but shifted progressively upward, indicating an independent relation between right ventricular free wall dimensions and left ventricular cavitary volume. An ellipsoidal shell subtraction model (pi/6.abd minus right ventricular free wall volume) was developed to estimate right ventricular cavitary volume from cardiac dimensions. With this method, a highly linear relation was observed between calculated right ventricular volume and right ventricular balloon volume (mean r = 0.99 +/- 0.01). Moreover, this relation appeared to be independent of changes in left ventricular balloon volume. With the shell subtraction model, dynamic right ventricular volume was computed in nine conscious dogs, and in four, stroke volume derived from dimensions was compared with right ventricular stroke volume measured with ultrasonic flow probes. A highly linear relation was observed, suggesting the accuracy of the shell subtraction method in vivo. Right ventricular end-systolic pressure-volume and stroke work/end-diastolic volume relations then were evaluated, and both proved to be highly linear in the right ventricle (both mean r = 0.99 +/- 0.01). Thus, the shell subtraction model allows a simple estimate of dynamic right ventricular volume in the intact heart and facilitates assessment of right ventricular performance in vivo.