This chapter explains counterfactual analysis as key for causal inference in impact evaluation, especially in education. It defines it as estimating what would have happened without an intervention, contrasting it with observed results. The chapter covers its theoretical basis, debates, and essential elements like comparison groups and assumptions that affect validity. It discusses experimental, quasi-experimental, and hybrid designs, detailing techniques such as t-tests, propensity score matching, and differences-in-differences, with an example from Digital Teachers for Education. This case shows how counterfactuals can test if digital skills improvements lead to better student engagement. The chapter also notes criticisms related to ethics, feasibility, epistemological limits, and transparency. Counterfactual analysis is a valuable tool that, used rigorously and alongside other methods, improves decision-making and learning in education.