Subgraph isomorphism problem on specialized graphs is an open research area. Most of the algorithms that had been developed for subgraph isomorphism problem are designed for general graphs. Although, they can be utilized for any graph type, this generic design prevents to take advantages of characteristics of the specialized graphs to burst performance of the subgraph isomorphism search. In this study, we will introduce a novel branch and bound algorithm for subgraph isomorphism problem on bipartite graphs. Bipartite feature of these graphs is utilized to improve search performance by limiting the size of possibility set and by getting rid of backtracking phase of branch and bound technique. Experimental results show that the algorithm provides between 10 and 120 fold better performance results on bipartite graphs when it is compared with the state-ofart algorithms.