A series of thickness of oxygen-enriched α layer were obtained after TC6 titanium alloy were treated with different thermal exposure processing (thermal exposure temperature, holding time). Their relationship network model was built by BP artificial neural network. The results show that the built model can be used for the prediction of the thickness of oxygen-enriched α layer of TC6 titanium alloy under different thermal exposure treatments. Meanwhile, the model can also serve as a guide for the heat treatment of TC6 titanium alloy if the thickness is given.