The influence of left ventricular filling and variations in end diastolic volume on cardiac performance was studied in the intact dog heart. Left ventricular filling volumes and stroke volumes were calculated on a beat to beat basis from measurements of natural mitral inflow and aortic outflow obtained by electromagnetic flow sensors. Instantaneous controlled modifications of end diastolic volume were performed through a cannula situated in the left ventricle and connected to a pump system outside the dog. This system enabled controlled increases or decreases of the end diastolic volume at any prechosen moment during the diastolic pause. Absolute volume variations in end diastolic volume and end systolic volumes could be calculated by combining the integrated flow signals from different consecutive beats. Left ventricular performance was evaluated in terms of end systolic volume and end systolic pressure variations. When the diastolic volume was abruptly increased by the pump system, natural mitral inflow decreased but end diastolic volume increased. The effect on diastolic pressure was dependent on the variation in filling rate, the amplitude of the infusion, the moment at which the infusion was started, and the diastolic pressure at the start of the infusion. Also stroke volume, maximal systolic pressure (Pmax), end systolic pressure, and end diastolic volume increased. The increased systolic performance was attributed to the increased end diastolic volume as expected according to Starling's law. When end diastolic volume was rapidly decreased during diastole by the pump, natural filling volumes increased to compensate for the volume loss by the pump. End diastolic volume was, however, smaller indicating that full compensation was not achieved. Evaluation of ventricular performance in terms of end systolic pressure and end systolic volume showed a decreased end systolic pressure and increased end systolic volume compared with the control values. The effect of a pump withdrawal was 1.62(0.38) times larger than could be explained on the basis of Starling's law. After the infusion of adrenaline the intrinsic depression disappeared and the influence of the volume withdrawal on cardiac performance was as expected from the Starling mechanism.