Far-field potentials have been produced from muscle tissue and shown to arise from both the proximal and distal musculotendinous junctions after the activation of a small group of muscle fibers. This investigation demonstrated that the polarity of muscle far-field potentials is consistent with the predictions of the leading/trailing dipole model. Far-field potential polarity was dependent upon the active electrode's orientation with respect to the positive or negative aspect of the trailing dipole after extinction of the leading dipole at the musculotendinous junction. This study also quantitatively measured the duration and magnitude of the two far-field potentials generated in the human biceps muscle by both proximal and distal muscle stimulation. The muscle far-field potential resulting from distal muscle stimulation closest to the musculotendinous junction was consistently shorter in mean duration (5.2msec +/- 2.1) but larger in mean amplitude (22.3 microV +/- 12.9) compared with the second far-field potential mean duration (10.4msec +/- 3.9) and amplitude (10.0 microV +/- 3.7). The mean areas under the curve for these far-field potentials, however, were comparable at 48.0msec microV and 48.7msec microV. Proximal muscle stimulation resulted in similar findings for the first far-field potential's mean duration, amplitude, and area (6.0msec; 19.7 microV; 50.6msec microV, respectively) in relation to those of the second far-field potential (12.5msec; 9.1 microV 49.1msec microV). The differences in amplitude and duration of the two far-field potentials, with similar areas, are consistent with, and can be adequately explained by, temporal dispersion effects of the muscle fiber action potentials' propagation over distance.