In the existing research on the three-stage assembly integrated scheduling problem model, the transportation time of each job is simplified to the same constant at the transportation stage. In practical situations, however, jobs are supposed to be transported separately in batches, due to the limitation of the number and the load capacity of transport vehicles. This paper studies the three-stage assembly integrated scheduling problem with job batches transportation (3sAISP_JBT) and the corresponding algorithm. The criterion of 3sAISP_JBT is to minimize the maximum completion time. Firstly, a mathematical model of 3sAISP_JBT is formulated in stages. Secondly, in order to reduce the overall computational complexity of solving 3sAISP_JBT, the first completed first transported(FCFT) rule and the first come first assembly(FCFA) rule are proposed for the sub-problems in the transportation and assembly stages, respectively. Thirdly, a hybrid estimation of distribution algorithm with rules(hybrid estimation of distribution algorithm with rules,HEDA_R) is proposed to solve 3sAISP_JBT. Finally, simulation experiments and algorithm comparisons verify the effectiveness of the proposed algorithms.