To solve unknown model parameters and the limitation of network communication resource of unmanned surface vehicles (USV), a data-driven event-triggered heading control method is presented for USV. An equivalent data model is constructed for the USV by applying input and output data of heading control system, and then the modelfree adaptive control scheme is utilized to construct data-driven heading controller. Since there is non-self-balancing characteristic in the heading control system, the historical rudder angle input compensation item is introduced into the rudder control input criteria to solve this problem. Furthermore, the event-triggered mechanism is constructed to decrease transmission signal of network and reduce the consumption of network communication resources by applying heading measurement error and tracking error. Finally, simulation results are implemented on the USV to verify the effectiveness of the designed data-driven event-triggered heading control method.