We used functional MRI (fMRI) and a network model based on graph theory to measure functional connectivity of brain motor network in the resting state in patients with Parkinson's disease (PD). FMRIs were acquired in 22 PD patients before and after levodopa administration, and in age- and sex-matched normal controls. The total connectivity degree of each region within the motor network was calculated and compared between patients and controls. We found that PD patients at off state had significantly decreased functional connectivity in the supplementary motor area, left dorsal lateral prefrontal cortex and left putamen, and had increased functional connectivity in the left cerebellum, left primary motor cortex, and left parietal cortex compared to normal subjects. Administration of levodopa relatively normalized the pattern of functional connectivity in PD patients. The functional connectivity in most of regions in the motor network correlated with the Unified Parkinson's Disease Rating Scale motor score in the patients. Our findings demonstrate that the pattern of functional connectivity of the motor network in the resting state is disrupted in PD. This change is secondary to dopamine deficiency, and related to the severity of the disease. We postulate that this abnormal functional connectivity of motor network in the baseline state is possibly an important factor contributing to some motor deficits in PD, e.g. akinesia.