This paper proposes a novelty heuristic search algorithm, called BSDBF (bigitem smallitem divide-and-conquer best-fit), to solve the two-dimensional rectangular fixed-size guillotine bin packing problem. First, based on group rules, this algorithm implements big item smalltime divide-and-conquer strategy and efficient group recommendation scheme which are key points to improve the group strategy. Then, the best-fit group is selected for recursive packing, and packing solution is achieved greedily for all bins. Finally, an initial solution is obtained, and a post processing algorithm is used to improve the quality of the solution based on item splitting method. That the solution can be obtained again is the critical characteristic of BSDBF algorithm which is different from others algorithms, because there is not any random factor in BSDBF algorithm. The computational results of many Benchmark problems have shown that BSDBF algorithm outperforms others reported algorithms.