This paper uses nationally representative data on high school students to test for several types of social influences on the decision to enroll in college. An instrumental variable strategy is used in order to manage the well-known reflection problem in social interactions research. Additionally, I am able to incorporate several usually unavailable group-level factors to reduce the possibility of important group-level characteristics driving the relationships. I present evidence that a 10 percentage point increase in the proportion of high school classmates who attend college is predicted to increase an individual's probability of attending college by approximately 2–3 percentage points. (JEL I2, J24, J18)