This work proposes using functional Near-Infrared Spectroscopy (fNIRS) as a non-invasive alternative to study the motor cortex’s functional connectivity in Parkinson’s Disease (PD). The bilateral motor regions were covered with the fNIRS probe, and graph theoretical network analysis and network-based statistics were applied to investigate differences in network topology and specific sub-networks between groups. Small-world properties like clustering coefficient, characteristic path length, and small-world index were computed and compared between PD patients and controls across various sparsity thresholds. PD patients exhibited a lower clustering coefficient and small-world index than controls. Network-based statistics identified a disconnected, mostly bilateral subnetwork in the PD group comprising nine edges and ten nodes. Mean functional connectivity was positively correlated with both groups' clustering coefficient and small world index, albeit this correlation was greater in the control group. A strong coupling between these two properties suggests that greater functional connectivity within the subnetwork may cause a more effective functional motor network in controls. The results provide insights into alterations in functional connectivity and network organization in the motor cortex of individuals with PD, demonstrating the potential of fNIRS for studying the neural basis of symptoms in this disease.