This paper estimates the contemporaneous effect of education on adolescent crime by exploiting the implementation a reform that increases the school leaving age in Italy by one year. We find that the Reform increases the enrollment rate of all ages, but decreases the offending rate of 14-year-olds only, who are the age group explicitly targeted by the Reform. The effect mainly comes from natives males, while females and immigrants are not affected. The Reform does not induce crime displacement in times of the year or of the day when the school is not in session, but it increases violent crimes at school. By using measures of enrollment and crime, as well data at the aggregate and individual level, this paper shows that compulsory education reforms have a crime reducing effect induced by incapacitation, but may also lead to an increase of crimes in school facilities plausibly due to a higher students concentration. JEL: I21, I28, J13, K42