As the information technologies are widely used in construction risk management, which can bring out explosive growth of data, how to organize and analyze the data becomes a crucial problem for construction managers. There are few issues to argue about the application of data mining in construction risk management. So, it is worthwhile to experiment in this particular field to manage the risks in construction projects. The major aim of this paper was to use data mining to extract implied, unknown, uncommon and potential application valuable information or patterns, which names Knowledge Discovery in Database. Therefore, correspondent effective countermeasures may be adopted in order to control the risk in construction project. First reviewed in this paper were the concepts of data mining and construction risk management. Several steps of risk management based on data mining were then shown, such as (1) identification of data mining object, (2) data preparation, (3) establishment of data mining model，(4) data mining，(5) result analysis，(6) knowledge application. Construction risk managers can apply the new knowledge with the help of data mining, thus making appropriate decisions faster, more pinpoint, and more comprehensive. Moreover, its possible applications can be extended to find effective measures in order to improve the level of other construction management.