In response to low take-up, many public schools have experimented with moving breakfast from the cafeteria to the classroom. We examine whether such a program increases performance as measured by standardized test scores, grades and attendance rates. We exploit quasi-random timing of program implementation that allows for a…