This study explored the feasibility of using functional NearInfrared Spectroscopy to confirm suspects in simulated theft cases. Using functional NearInfrared Spectroscopy technology, a simulated theft case experiment was designed. Thirty volunteers were selected and divided into a criminal group, an informed group, and an innocent group. The experiment simulated a real interrogation scenario of a crime, using Guilty Knowledge Test’s method to develop questions. The brain activation levels and blood oxygen concentration changes of the suspects were measured when they lied and answered honestly. The study found that the spectral maps of the three groups of subjects all had unique characteristics. The criminal group showed significant activation in the prefrontal lobe and left temporal lobe regions, accompanied by an increase in blood oxygen concentration; The informed group showed partial activation of brain regions and an increase in blood oxygen content concentration; The brain regions of the innocent group were not activated, and the blood oxygen concentration tended to be stable without significant fluctuations. Based on this characteristic, near-infrared technology can accurately identify the suspect in a simulated theft case, demonstrating high accuracy and stability, providing strong support for the police in determining the suspect’s identity.