Using the theory and method of artificial neural network, this paper established a decision-making model of bid evaluation for engineering projects based on back propagation neural network (BPNN). BP neural network was trained and tested by the learning sample that was attained by using order relation analysis method (G 1 -method). At the same time, we used the momentum BP algorithm to solve the model. The results show that the model is feasible and reasonable, possessing high abilities of self-organization, self-adaptation and self-learning and strong function of fault-tolerance, avoiding weaknesses of slow convergence speed, easily falling into local minimum and frequent occurrence of oscillation in learning process of traditional algorithms, eliminating bias caused by arbitrary individual experts, improving efficiency of bid evaluation, and providing effective reference and basis for bid evaluation activities of general engineering projects.