Eight conscious dogs instrumented with wall thickness sonomicrometers and 11 subcutaneous electrodes in a modified McFee vectorcardiographic array were studied during changes in ventricular volume. Simultaneous measurements were made of QRS amplitudes of the endocardial and epicardial ECG, QRS spatial vector magnitudes (SVM), end-diastolic wall thickness (EDT), end-systolic wall thickness (EST) and the amount of systolic thickening (zAWT). Ventricular size was decreased by atropine and infulsion of 0.02 gg/kg/min of isoproterenol to increase the mean heart rate from 81 ± 5 beats/min (mean ± SEM) to 174 10 beats/min (p < 0.001), and was reflected by an increased mean EDT (9.06 ± 0.64 mm to 9.94 i 0.61 mm, p < 0.005). The endocardial QRS amplitude increased in each dog (mean increase 21.55 ± 1.36 mV to 25.13 ± 1.35 mV, p < 0.001), whereas the SVM decreased from 7.69 ± 0.75 mV to 6.18 ± 0.48 mV (p < 0.02). Ventricular size was then increased by rapid saline infusion and was reflected by a decrease of EDT from 9.65 ± 0.66 mm to 9.09 ± 0.66 mm (p < 0.001), while heart rate remained unchanged. Endocardial amplitude decreased in each dog (average decrease 3.59 ± 0.25 mV, p < 0.001), while the SVM increased in each dog (average increase 0.81 ± 0.18 mV, p < 0.005). The mean epicardial amplitudes did not change significantly during either increases or decreases in ventricular volume. In each dog, there was a linear relation between EDT and endocardial amplitudes (r values > 0.88) and an inverse linear relation between EDT and SVM (r values > -0.80). The relations between EST or AWT and regional and QRS surface amplitudes were nonlinear. We conclude that in the conscious dog changes in endocardial QRS amplitudes and SVM respond in an opposite manner to changes in ventricular volume. In this experimental model, alterations in endocardial QRS amplitudes were related directly to changes in diastolic wall thickness; changes in body surface QRS amplitudes were inversely related to wall thickness, a finding that may relate in part to alterations in the distance of the heart from the chest wall. THE MECHANISMS underlying regional and surface QRS amplitude changes during acute forms of stress are unclear, and it has not been established whether changes in left ventricular volume are directly or inversely related to QRS amplitudes. In model systems 14 and in animal experiments,5' 6 an increase in ventricular volume produces an augmentation of the initial portions of the QRS potentials in the body surface ECG and vectorcardiogram (VCG), but in human subjects the direction of such changes has been controversial.7' 8 Lekven et al. recently reported that the endocardial QRS amplitude decreased during volume increases in open-chest dogs,9 an effect opposite to that described in the surface ECG.5' 6 It has also been proposed that changes in the ventricular ejection fraction relate to surface QRS-amplitude changes.10 11 However, clear relations between left ventricular ejection fractions, left ventricular volumes and alterations in surface...