It is difficult to explain drilling conditions in a short time with the increased application of MWD tools and the increased data volume and the density of inherent data during the drilling. How to use these data to efficiently guide drilling operation and reduce drilling risk becomes a problem which puzzles drilling engineer. During the drilling, case-based reasoning can make a good interfusion of the acquisition data and the real-time analysis data and simulate the cognitive processes of mankind to make reasoning analysis of risk which is come across and timely and correctly predicted, and the various complex issues are solved during the process of drilling. By studying the principle of case-based reasoning and taking the lost circulation risk as an example, this paper elaborates the risk predication process based on case-based reasoning, and expounds the technologic process and application example of DrillEdge system. The results of study show that this method is simple and practical, which can predict drilling risk quickly and accurately and can provide good supports on site.