Research and optimizing on laser strengthening processing and controlling for processing 20CrMo alloy structure steel by taking laser power, laser processing beam diameter, laser scanning velocity as input parameters and non-transformation hardening, transformation hardening, melting as output parameters. BP neural network of processing optimal model between laser processing parameters and material surface strengthened quality is established by comparing experimental results and computering simulation analytic results. It is shown that laser processing parameters can be chosen conveniently and material surface quality can be controlled effectively by using BP neural network, furthermore, the proposed model provides accurate results for processing, and fits the experimental results well.