Recently, with the rapid development of big data, the accuracy, speed, efficiency and the cost of remote sensing yield estimation have been greatly improved. Based on the relevant literature researches of domestic and foreign scholars in the remote sensing yield estimation for nearly ten years, we divide the process into four stages: the crop planting area extraction, the crop growth monitoring, the crop yield estimation model, and the model accuracy evaluation. In this paper, we also conclude the characteristics of each stage, evaluate the advantages and disadvantages of the commonly used methods in each stage, and explore solutions to some main problems in the field of remote sensing yield estimation, finally, through analysis and summary, we give some feasible suggestions to use the models in different situations to provide references for the future research.