Natural experiments have become increasingly popular in the social sciences. In this article I give an overview of the concept. I discuss its definition and origins, the reasons why researchers increasingly rely on natural experiments, and the ways to evaluate natural experimental designs. Next to several examples, the article includes a hands-on application with open data and code. The application goes through all stages of using a natural experiment, by describing the case, evaluating the design by placebo tests and diagnostics, as well as presenting and interpreting the results. The discussion also includes tips in finding natural experiments with a note on if it was always necessary to find one, and what to do if none could be found. Potential pitfalls of natural experiments are also discussed. Overall, natural experimental designs, when they are used responsibly and carefully, strongly enrich the methodological arsenal of social scientists.