OBJECTIVE: To develop a method to place a lesion precisely in the subthalamic nucleus (STN) and evaluate its effectiveness. METHODS: A retrospective study of targeting data collected during stereotactic planning to lesion the STN in 31 patients with Parkinson's disease and of results in more than 50 procedures was performed. The targeting method was based on computed tomographic imaging together with semimicroelectrode recording digital processing and electrical stimulation. Two statistical methods were used to correlate initial with final target coordinates and assess the efficacy of the targeting procedure. RESULTS: The anatomic target based on computed tomographic imaging data showed electrical activity in the subthalamus in the first pass in 82% of the procedures. In the remaining 18%, the STN was an average of 1.93 mm away from the nearest trajectory that recorded the STN (range, 1.41-2.24 mm). The average number of trajectories per procedure was 7.2; the location of the first trajectory relative to the center of the nucleus determined by electrical and physiological means (P < 0.01, analysis of variance, Student's t test) was as follows: in the lateral direction, 1.25 +/- 1.15 mm; in the anteroposterior direction, 1.53 +/- 1.31 mm; and in the vertical direction, 0.67 +/- 0.51 mm. The average number of tracts necessary to lesion the STN was two. CONCLUSION: The combination of computed tomographic imaging, semimicroelectrode recording, and microstimulation provides an effective method to identify the STN lesion in parkinsonian patients. The method used for anatomic localization and electrophysiological mapping of the subthalamus was found to be effective in reaching the sensorimotor region of the nucleus. We carried out an accurate determination of the subthalamus location and its volume in the lesioning.