Intraoperative subcortical electrical stimulation is used to identify and preserve white matter tracts so that tumor resection can be performed while avoiding postsurgical deficits. The effects of the stimulating electrodes in identifying the white matter tracts have not been characterized; thus, different hospitals use different electrode configurations. Computational modeling can be used to conduct a systematic assessment of the effects of the stimulating electrode parameters. However, no realistic computational model of subcortical electrical stimulation has been implemented and verified. In this study, we investigated the interaction between the corticospinal tract (CST) and subcortical stimulation and compared different electrode configurations during monopolar and bipolar stimulation. For that, we computed the induced electric field in a realistic human head model coupled with a CST axon model. The implemented model was verified with available experimental data that were acquired during subcortical stimulation, and a systematic sensitivity analysis of parameters related to the stimulation was conducted. The results showed that the optimal stimulation varies according to the surgery conditions. If the CST was close to the resection border, bipolar stimulation could produce more selective activation. Monopolar stimulation was more robust and more effective for the CST far from the stimulation point.