Abstract: For on the application characteristics of BP (back propagation) neural network in the investment estimation of engineering projects, combined with the characteristics of investment estimation of prefabricated buildings, the model of BP neural network based on beetle antennae search (BAS) algorithm was put forward. The investment estimation of typical cases of 15 assembly buildings was analyzed by BP neural network model and the proposed BAS-BP network model.The results show that the prediction accuracy of BAS-BP network model is better than that of BP neural network model, and the prediction results are more stable, which is more suitable for the investment estimation of prefabricated buildings.