Economists have a poor understanding of the mechanisms underlying reduced-form college peer effects. In this paper we explore a candidate mechanism, the provision of school effort. We show that, when earnings reflect individual educational performance as well as the field of study selected at college, and individual effort is a function of expected earnings, the size of the peer effect varies by field. Using data from a middle-sized public university located in Southern Italy and exploiting the random assignment of first year students to college accommodation, we find evidence that peer effects are positive and statistically significant for students enrolled in the fields of Engineering, Maths and Natural Sciences – which are expected to generate higher earnings after college – and not different from zero for students enrolled in the Humanities, Social and Life Sciences, which give access to lower payoffs. An implication of our model is that shocks affecting college wage premia may alter the size of peer effects.