In terms of complex electronic systems, in order to enhance the test efficiency and reduce the test cost, an optimization method of analog circuit test point combining with fault dictionary and branch and bound is proposed. It adopts fuzzy set and fault dictionary as tools, and takes fault detection and fault isolation as constraint condition, thus to build a 0-1 programming mathematical model for test point optimization, which is solved with branch and bound. Finally, it simulates and verifies the method with classical test optimization circuit. The simulation result indicates that the above method can realize rapid optimization of test point on the premise of guaranteeing the demand of fault diagnosis, providing guidance for test implementation.