In view of the complicated, changeful and uncertain air-combat environment, the fuzzy Petri net based reasoning algorithm was studied. Then, an intelligent tactical decision-making method based on fuzzy reasoning Petri net was established, which changed the qualitative decision reasoning process into quantitative matrix operation. For the redundancy and possible conflict of rule, rough set theory was used to refine the tactical decision-making rule database, thus to simplify the fuzzy reasoning Petri net structure, improve the efficiency of tactical decision reasoning, as well as realize intelligent decision-making for UAV attack and defence. Simulation results demonstrate the feasibility and validity of the proposed method