The use of project portfolio management is increasingly becoming a tool for promoting the strategy of the organization. Using sophisticated quantitative tools becomes a significant competitive advantage for project portfolio management. Project portfolio management is a dynamic multi-criteria decision-making problem under risk. The paper presents new approaches for analyzing the problem. A dynamic version of the Analytic Network Process (ANP) captures the network, multicriteria and dynamic structure of the problem. Multicriteria decision trees analyze risk of project portfolios. Possible projects are characterized by sets of inputs and outputs, where inputs are resources for project realization and outputs measure multiple criteria of goals of the organization. The Data Envelopment Analysis (DEA) is an appropriate approach to select efficient project portfolios.