Plant breeding is the art of genetic improvement through creation and selection for novel characteristics in plants. Parental selection provides the raw materials for creating each new generation of genetic improvements. Ultimately, the probability of successfully meeting the breeding objectives depends on selection of parents for intermating. With the application of operations research, we developed a new metric, parental breeding value (PBV), based on application of conditional probability distribution to help to solve the parental selection problem to accelerate the process of plant breeding and save resources at the same time. The water pipe model is provided in the thesis to calculate PBV efficiently. Also, we discuss the potential of Markov Decision Processes to address the challenge. For small scale cases, the MDP based approach will lead to a precise result with better performance. However, when dealing with large scale problem, it will suffer calculation complexity and can hardly be applied to realistic problems. In order to fix this, we try simulations based on PBV which can overcome the limitation of the MDP approach. From the results of simulations, the PBV metric is demonstrated to shorten the plant breeding process and decrease the resources costs. We can conclude that the PBV will contribute to deriving better strategies for realistic plant breeding objectives.