Abstract: Current hydropower development in the lower Jinsha River encounters great technical and management challenges, including complicated and harsh natural conditions, rapid changes in social environments, diversified interests, and trans-regional multi sub-projects. This requires further improvement on the existing capability of large-scale project construction for forecast, analysis and control of the projects to minimize their uncertainties brought about by human factors, changing environments and construction conditions, so that the target of six managements of construction projects can be better achieved. To this end, a concept of intelligent dam was formulated and introduced into the construction management of Xiluodu ultra-high arch dam, and intelligent construction technologies were developed and applied to typical dam construction procedures. Based on these applications, this paper describes a theory of intelligent management of large-scale construction projects that was formulated by combining current project management trends with extensive applications of information technology. We also present a technical framework of this new type of management that applies a closed-loop intelligent control principle developed in this study and featured with comprehensive sense, accurate analysis, and real-time control. This framework have five key components: starting from a 3D GIS/DIM structure and multi-source data, targeting at the capability of creating life-long asset value, dynamic simulation and real-time control of actual working states as its core technology, a TGPMS/iDam system as a platform for comprehensive task integration, and N-dimensional enhanced collaborative management as its guideline. The new management framework and theory presented in this paper would be helpful in shaping and upgrading traditional construction project management to a new level characterized by informatization, digitalization, and intellectualization to achieve better project construction in quality, safety and efficiency.