The uncertainties and randomness in multi-dimensional risk management (RM) in large-scale construction projects should be effectively detected and measured through reliable managerial procedures to eliminate or reduce adverse consequences such as casualties and asset damages in advance. This paper aims at an innovative technical solution for project managers to better understand the level of RM practice by means of RM maturity measurement. Based on the theory of system science and previous research into project management in the maturity of RM, this paper describes the prototype of an RM Maturity System (RMMS) for large-scale construction projects. The system is underpinned by using the Analytic Network Process (ANP) method in order to measure the overall effectiveness of RM against major risk factors. The RMMS consists of three components to focus on capabilities, evaluation and evolution in RM. A case study is given at the end to verify the feasibility of the system.