With the development of automobile technology, the function of engine detection device is more and more obvious. The introduction of virtual instrument can build a general testing system with simple hardware. The purpose of this paper is to study engine fault diagnosis with virtual instrument technology. In view of the fact that the electronic and integrated level of automobile products is constantly improving, the basic theory and technical methods of fault self-diagnosis of automobile electronic injection system are studied. BP neural network theory and virtual instrument technology are comprehensively applied to the field of engine fault self-diagnosis. By analyzing the characteristic parameters containing rich engine status information, fault characteristics are extracted, A general platform of engine intelligent fault diagnosis based on virtual instrument technology is proposed and designed. The experimental results show that the system improves the diagnosis speed and accuracy.