The influence of nonuniform cell shape and field orientation on the field stimulation thresholds of cardiac myocytes was studied both experimentally and computationally. The percent change in excitation threshold, which was studied with patch clamp technique, was found to be 182 +/- 83.1% (mean +/- SD) higher when the electric field (EF) was parallel to the transverse cell axis versus the longitudinal axis (p < 0.0006). On reversing the polarity of the applied EF, the percentage change in threshold was observed to increase by 98.9 +/- 71.0% (p < 0.0002), implying asymmetry of the stimulation threshold of isolated myocytes. Finite element models were made to investigate the distribution of the transmembrane potential of these experimentally studied myocytes. A typical cell model showed that the maximum transmembrane potential induced on opposite ends of the cell was 39.1 mV and -46.5 mV for longitudinal field (aligned with the long axis of the cell), but was 40.5 mV and -44.8 mV for transverse field (aligned with the short axis of the cell). More significantly, it was found that the maximum transmembrane potential occurred at discrete points or "hot spots" on the cell membrane. It is hypothesized that the depolarization of the cell initiates at the hot spot and then spreads over the entire cell. The different excitation thresholds for different polarities of the applied EF can be explained by the different maximum induced at the opposite ends of the cell.