To facilitate the analysis of aortic-ventricular coupling, simplified models of aortic input properties have been developed, such as the three-element Windkessel. Even though the impedance spectrum of the Windkessel reproduces the gross features of the real aortic input impedance, it fails to reproduce many of its details. In the present study we assessed the physiological significance of the differences between real and Windkessel impedance. We measured aortic input impedance spectra from five anesthetized open-chest dogs under a wide range of conditions. For each experimentally determined spectrum we estimated the corresponding values of the best-fit Windkessel parameters. By computer simulation we imposed both the real and best-fit Windkessel impedances on a model left ventricle and assessed the differences in seven different coupling variables. The analysis indicated that the Windkessel model provides a reasonable representation of afterload for purposes of predicting stroke volume, stroke work, oxygen consumption, and systolic and diastolic aortic pressures. However, the Windkessel model significantly underestimates peak aortic flow, slightly underestimates mean arterial pressure, and, of course, does not provide realistic aortic pressure and flow waveforms.