The work aims to use finite element simulation to predict the corrosion site and corrosion rate of titanium-steel lap joints for aircraft. Polarization curves of TC18 titanium alloy, 30CrMnSiNi2A high strength steel and 30CrMnSiNi2A galvanized steel were measured by potentiodynamic polarization in 5%NaCl solution at 40 ℃ and with pH=4.0. The simulation model of galvanic corrosion for lap joints was established with the polarization curve and its fitted electrochemical kinetic parameters as boundary conditions. The surface potential and local current density distribution under perfect and completely damaged state of galvanized layer were obtained through the model calculation, and the simulation results were compared with the experimental results. The corrosion occurred on the surface of galvanized steel fasteners preferentially at the early stage. Corrosion of zinc layer started from pitting corrosion and expanded slowly, and gradually developed from local corrosion to uniform corrosion. After five cycles, the zinc layer was completely broken and 30CrMnSiNi2A of fastener was exposed and changed from the cathodic polarity to the anode polarity. 30CrMnSiNi2A and TC18 formed a new couple pair and corrosion expanded rapidly on the surface of the steel plate. The corrosion morphology of lap area after third cycle and seventh cycle corrosion test was compared with the results of simulated potential distribution of intact galvanized layer and completely damaged zinc coating. The corrosion morphology was basically similar to that of low potential region of potential distribution. According to Faraday‘s law and current density distribution, the corrosion depth of corroded area was predicted and compared with the result of corrosion test after third cycle and seventh cycle corrosion test, the predicted corrosion depth of simulation model is agreement with the result of corrosion test after third cycle and seventh cycle corrosion. The experimental results are consistent with the simulation results, thus proving the correctness of the simulation model.