Aiming at the decision of unmanned combat aerial vehicles （UCAVs） autonomous maneuver in 1vs.1 aerial combats， a decision method based on fuzzy inference is proposed. Firstly， four advantage factors of angle， distance， speed， and height are established from the air combat scene. Secondly， fuzzy rules are used to evaluate the air combat situation. The advantage factors representing the air combat situation are input into the fuzzy inference machine， and the weight of each factor in the advantage function is adjusted adaptively. At the same time， the enemy aircraft position prediction model is established by the combination of decision maneuvers， sequence maneuvers and inertial maneuvers， as an auxiliary means of maneuver decision-making， and finally uses trial maneuvers to make maneuver decisions. The simulation results show that the fuzzy inference model can realize UCAV autonomous maneuver decision ability. The advantage of the model lies in transforming the fuzzy air battle situation into the determined weight factor in the advantage function and guiding UCAV to make more targeted maneuver decisions strategically.