INTRODUCTION: Deep brain stimulation (DBS) is used to treat advanced Parkinson's disease (PD). Stimulation of the subthalamic nucleus (STN) is effective to improve the symptoms related to parkinsonism, but its mechanism of action remains poorly understood. Functional magnetic resonance imaging (fMRI) can be applied to study brain areas involved in motor activity, as a mean to better understand the effects of DBS. In this work we aimed to develop and test fMRI techniques to study DP patients treated with DBS. METHODS: We have initially performed tests to check the interference of the DBS in image quality and made adaptations that minimized the artifacts. After this initial phase PD patients were examined by fMRI in four different phases: during right hand movement (before surgery to implant DBS electrodes, after the surgery with the electrical stimulation turned off, and after the surgery with the electrical stimulation turned on the STN) and at rest with electrical stimulation turned on. In total, ten patients were examined, eight of them paired to matched normal volunteers. All tasks were performed in three independent sessions for each condition, using two types of runs: blocked (BK) and event-related (ER). The images were analyzed calculating the frequency of activation detected in 19 sensory-motor areas. Behavioral data during fMRI exams and clinical data were also recorded. RESULTS: There were no clinical complications due to fMRI exams. The modifications in the DBS procedure and post-surgical time to fMRI examination resulted in a reduction of the image artifacts. In the controls the ER runs showed more activity than BK in: supplementary motor area, right pre-motor area (pM), anterior cingulated gyrus (aCg), posterior cingulated gyrus (pCg), bilateral thalamus, and right putamen. During the dominant hand movements, when comparing ER and BK runs in PD patients there was no significant difference considering pre and post operative phases without electrical stimulation. When the electrical stimulation was turned on, BK runs showed more activation in the left primary sensory-motor cortex (SM1) compared to ER runs. In the rest state, when the electrical stimulation was turned on, there was more activation in the following areas: right cerebellum, left SM1, pCg, bilateral pM, left mesencephalus and aCg. There was clinical improvement in the patients, and the motor performance in the first post operative week, in the drug withdrawn phase as well as in the global evaluation after six months were correlated with the motor parameters recorded during the fMRI exams. CONCLUSIONS: we have developed a method for fMRI in PD patients with DBS, which enabled the following observations: BK runs showed increased frequency of activity in SM1 in PD patients with electrical stimulation turned on, while the ER runs showed more frequency of activation in normal volunteers in primary and secondary motor regions; the electrical stimulations in rest state produced activity detected in cerebellum, primary and secondary motor regions.