We often want to know if one thing causes another. Does more education raise wages? Does a new drug lower blood pressure? But the world is messy. Many other factors get in the way. This chapter shows how adding control variables to a regression can help us find a causal effect, if we do it right. We will learn the key assumption that makes this possible. We will also see a neat trick that shows what the regression is really doing. And we will learn how to pick controls wisely.