The objective of this project is to enhance knowledge in the construction, identification and estimation of dynamic structural microeconomic models that are used for policy evaluation. This research proposal is built up having specific economic applications in mind as these applications involve inter-temporal trade-offs for a single or several decision makers. It first seeks to develop original identification results in these applications and attaches special attention to partial identification issues and constructive identification results so as to easily derive estimation techniques. In each specific application, empirical estimates using micro-data will then be used to construct and analyse counterfactuals. The whole sequence of original identification, estimation and prediction results aims at enhancing the quality and credibility of economic policy evaluations.These research questions will be addressed in frameworks in which dynamic choices are continuous such as the ones regarding human capital investments or discrete such as the college choice decisions. This extends to dynamic games as in the analysis of firms' entry into a market.This research proposal develops micro-econometric analyses devoted to earning dynamics, consumption smoothing and incomplete markets, firms' entry, school matching mechanisms as well as to the dynamics of undergraduate studies and the dynamics of retirement. It involves studies in labor economics, consumer behavior as well as financial econometrics, empirical industrial organization and the economics of education. One last theme of this project is devoted to research in theoretical econometrics analyzing questions derived from the empirical projects. Each empirical project will cross fertilize others and will feed up theoretical econometric analyses related to point or partial identification in various dimensions. In turn, theoretical analyses will inform identification and estimation in each of those specific applications.