Aiming at the problem of multiple UAVs confliction resolution, a scheme based on modified artificial bee colony (MABC) algorithm was proposed. On the basis of the artificial bee colony (ABC) algorithm, the probability function and searching mechanism for the onlooker bees are modified to make the best solution play the role of guidance to enhance the exploitation capability and improve the convergence performance. The abilities for exploration and jumping out from local optimum are kept. Both of heading change resolution and speed change resolution are utilized to solve the confliction. The results demonstrate good performance for the convergence speed, running speed and the fitness value of best solution when compared with genetic algorithm.