The explanation of the response to electrical stimulus by macroscopic regions of cardiac tissue in terms of the behavior of membrane ion channels requires mathematical models that span the range of spatial scales from the single cardiac cell to the entire heart. This is accomplished by the bidomain model, which successfully characterizes the electrical properties of the heart and the effect of externally applied electric fields on myocardial tissue, Recently the bidomain model has been used to make several specific, testable predictions: (1) a 4-fold symmetric magnetic field pattern is associated with an expanding wave front, (2) a region of positive interstitial potential precedes an expanding wave front in the direction parallel to the myocardial fibers, (3) the rate of rise of an action potential depends on the direction of propagation in superfused tissue, (4) the wave front in superfused strands of tissue is curved (5) a "dog bone" shaped region of depolarization exists under a unipolar cathode, (6) depolarized regions along the fiber direction adjacent to a unipolar anode are responsible for anodal stimulation, (7) interactions between adjacent depolarized and hyperpolarized tissue cause anode-break and cathode-break stimulation, (8) reentry can be induced by successive stimulation using a single unipolar cathode, and (9) a mechanism for far field stimulation depends on fiber curvature. These predictions have been verified qualitatively in every case where they have been tested experimentally. In some cases, such as the mechanisms for reentry induction and far field stimulation, the necessary experiments have not yet been performed.