Incidents of violence have intensified concern about student conduct in America’s schools, leading to a desire for greater insight into its causes. This study investigates adolescent development in schools by examining the impact of family, peer and school traits on student conduct, with a particular interest in the type of high school attended. It advances prior research by Anthony Bryk and James Coleman by employing extensive controls for family traits. The findings indicate that influences of Catholic schools on student conduct are different than claimed in prior research. The study also investigates the dynamics of class lesson difficulty, a neglected factor in school reform. Through the use of a large national sample this work provides insight into adolescent development, school type and education reform.