Impedance matching in the cardiovascular system is discussed in light of two models of ventricle and load: a Thevenin equivalent consisting of a hydromotive pressure source and an internal, source resistance and compliance in parallel; and a time-varying compliance filled from a constant pressure source and ejecting into a load of three components, a central resistor, a compliance, and a peripheral resistance. According to the Thevenin analog, the energy source and the load are matched when the load resistance is T/t times the internal source resistance (T is total cycle length, t is systolic time interval). Both from this model and from the variable compliance model it appears that optimum matching between source and load depends on the compliance of the Windkessel, as low compliance shifts the matching load resistance to a low value. Animal experiments (isolated cat hearts) indicated that both left and right ventricles at normal loads work close to their maxima of output hydraulic power, and, according to experiments in the right ventricle, maximum power output is related to load resistance and compliance as predicted by the above models. From an experimentally determined relationship among instantaneous ventricular pressure and volume (right ventricle of isolated cat hearts), an optimum load impedance was calculated on the basis of the assumption that the ratio between stroke work and static, potential energy developed in the ventricular cavity is maximum. The optimum load impedance found by this procedure closely resembles the normal input impedance of the cat lung vessel bed.