This paper focuses on how to optimize cascaded hydropower investment. First, we analyze why the traditional optimization methods are not suitable for cascaded hydropower investment optimization, and then we discuss the dynamic programming approach, finally, we construct a dynamic programming model for cascaded hydropower investment optimization. The research shows that traditional methods, such as NPV, Real Option and Game Theory, mostly are only suitable for investment decision-making analysis of the single-period projects, but not appropriate for the investment optimization of multi-period projects. The cascaded hydropower investment is a typical question with multi-period optimization, and the theoretical anylisis and calculation example both demonstrate the dynamic programming is feasible and efficient for it. The paper recommends that dynamic programming should be adopted to improve the investment efficiency of cascaded hydropower project.