Estimation of Distribution Algorithm (EDA) is a new kinds of colony evolution algorithm, through counting excellent information of individuals of the present colony EDA construct probability distribution model, then sample the model produces newt generation. To solve the NP-Hard question as EDA searching optimum network structure a new Maximum Entropy Distribution Algorithm (MEEDA) is provided. The algorithm takes Jaynes principle as the basis, makes use of the maximum entropy of random variables to estimate the minimum bias probability distribution of random variables, and then regard it as the evolution model of the algorithm, which produces the optimal/near optimal solution. Then this paper presents a rough programming model for job shop scheduling under uncertain information problem. The method overcomes the defects of traditional methods which need pre-set authorized characteristics or amount described attributes, designs multi-objective optimization mechanism and expands the application space of a rough set in the issue of job shop scheduling under uncertain information environment. Due to the complexity of the proposed model, traditional algorithms have low capability in producing a feasible solution. We use MEEDA in order to enable a definition of a solution within a reasonable amount of time. We assume that machine flexibility in processing operations to decrease the complexity of the proposed model. Muth and Thompson’s benchmark problems tests are used to verify and validate the proposed rough programming model and its algorithm. The computational results obtained by MEEDA are compared with GA. The compared results prove the effectiveness of MEEDA in the job shop scheduling problem under uncertain information environment.