Abstract: Grain production is the important part of national economy development, whose fluctuation will make much effect on national economy safety. Therefore, people also need to know the grain production changes ahead of crop harvest period while they make an effort to increase grain production, for the purpose of providing the basis of scientific decision-making. Based on the grain yield data from 1949 to 2008 of Jilin province, Grey Prediction Model (1, 1) was utilized for dynamic simulation of the yield changes, and Markova Transition Matrix was introduced to amend the simulated result to improve the yield predicted accuracy. The studied results indicated that, the accuracy of the amended simulating yield increased greatly compared with that simulated by original GM (1, 1) (the relative error decreased obviously from 0.19 to 0.09). It is feasible of using a combination of grey model GM (1, 1) and Markov Chain Method for simulated grain yield. It was found that the total grain yield of Jilin Province will increase 10 billion kilogram in the next 10 years, which showed the great potential of grain producing in Jilin province.