Two independent mathematical models of the electrophysiology of the human atrial cell have been published. Although these two models are based on very similar data, they can result in some apparently contradictory findings, some of which are of fundamental importance to human atrial electrophysiology and pharmacology. In particular, simulations with these two models have given rise to different conclusions regarding the role of the rapidly activating sustained outward K + current, I sus , in determining the action–potential duration. Different conclusions regarding the rate dependence of the action potential have also been published. In this paper, we review and compare these two models in an attempt to determine the reasons for these discrepancies. The individual ion current characterizations are quite similar in the two models. The primary difference between the models is in the assumed baseline actionpotential shape, and therefore the sizes of the underlying ionic currents. The duration of spike–and–dome–type action potentials, which rely mainly on the delayed rectifier currents, I Kr and I Ks , for final repolarization, is relatively insensitive to block of I sus . In contrast, r triangu s lar action potentials are prolonged substantially by I sus block, since they rely more on I sus for final repolarization. The rate–dependent properties of spike–and–dome action potentials can be explained in terms of their larger I CaL (L–type calcium current), I Kr and I Ks currents.