A physiologically based multicompartmental computational model of a midbrain dopamine (DA) neuron, calibrated using data from the literature, was developed and used to test the hypothesis that sodium dynamics drive the generation of a slow oscillation postulated to underlie NMDA-evoked bursting activity in a slice preparation. The full compartmental model was reduced to three compartments and ultimately to two variables, while retaining the biophysical interpretation of all parameters. A phase-plane analysis then suggested two mechanisms for the regulation of the firing pattern: (1) bursting activity is favored by manipulations that enhance the region of negative slope in the whole-cell IV curve and inhibited by those manipulations, such as increasing linear currents, that tend to dampen this region and (2) assuming a region of negative slope is present in the IV curve, the bias of the system can be altered, either enabling or disabling bursting. The model provides a coherent framework for interpreting the effects of glutamate, aspartate, NMDA, and GABA agonists and antagonists under current-clamp conditions, as well as the effects of NMDA and barium under voltage-clamp conditions.