After a careful review of academic works that apply the real options approach to the evaluation of infrastructure concession projects and identifying the main gaps in this literature, we developed this Thesis, composed of four independent studies. The first explains the binomial model and shows how to incorporate the project cash flows using the cash flow dividend rate to create the project value lattice. We develop a R code, provide a tutorial on how to use this model and show how the code can be customized for particular applications. The second shows why additional investments in expansion as firm obligations in concession contracts are suboptimal and proposes a real options model that combines flexible capacity expansion decisions with conditional term extensions. Using a typical toll road project in Brazil, we show how this kind of flexibility can be useful for policy development to attract private investment in public infrastructure projects. The third evaluates the concession of a Light Rail Vehicle in Brazil. We adopt the real options approach to model the different flexible clauses embedded in this contract and analyze whether they conflict with each other and how they impact the overall project evaluation. Finally, the fourth uses Unit Root, Variance Ratio tests, and the Parameter Approach Measure to evaluate the most appropriate stochastic process to model the uncertainty of passenger demand in airport concessions in Brazil. We analyze samples ex-ante and ex-post covid-19 and show that both seasonality and the pandemic significantly impact the stochastic diffusion model.