Multiple UCAVs air-to-air combat poses great challenges such as a vast number of input variables, complex confrontation rules that are difficult to be obtained, etc. A tactical decision method based on single input rule modules (SIRMs) dynamically connected fuzzy inference model and an improved adaptive genetic algorithm (IAGA) for multiple UCAVs air-to-air combat is proposed in this paper. Considering the “dimension disaster” problem in traditional fuzzy rules design due to increasing input variables, SIRMs dynamically connected fuzzy inference model is used to make tactical decision, where all input variables are decoupled by SIRMs and the number of rules is greatly reduced. The model output is merged by adding the results of all decoupled rules with dynamic importance degrees. As accurate confrontation rules are difficult to be obtained in practice, the IAGA is designed to optimize the consequent part of the rules, which can produce detailed rules using only a simple rule skeleton. Two representative 2 vs 1 cases are simulated. The results show the validity, universality, and expansibility of the tactical decision method proposed in this paper. It also can be easily extended to n vs n cases in different combat situations.