The function of the pressure-overload left ventricle was studied in relation to the chamber geometry of the isolated canine heart. The occurrence of concentric hypertrophy was confirmed in dogs with aortic constriction 40 weeks after operation. The effect of the concentric hypertrophy on left ventricular function was then studied in five pressure-overload dogs. The end-diastolic pressure of the preparation was preset at 20 mmHg. In the pressure-overload ventricle, the end-diastolic volume was smaller by 47% (p less than 0.01) and the peak systolic pressure was higher by 24% (p less than 0.01) than the control at isovolumic beats. Ejection pressure was controlled by our afterload controlling system at various levels and kept constant during the ejection phase. When ejection pressure was the same, the difference in stroke volume between the experimental and control groups was not significant. In the pressure-overload ventricle, the slope (the reciprocal is Emax) of the regression line of the end-systolic volume on ejection pressure decreased (p less than 0.01) to 0.095 ml/mmHg by 47% from 0.186 ml/mmHg in the control. The volume-axis intercept (Vd) of the line was reduced to 55% of the control. A decrease in the regression line slope was in linear proportion to the degree of concentric hypertrophy, namely, to the change in chamber geometry. These results indicated that: Despite reduction of the end-diastolic volume, the concentric hypertrophied ventricle showed no reduction in stroke volume at the same level of ejection pressure, because of its improved capability to generate pressure. Emax was dependent on ventricular geometry.