No nationally representative study of student suspensions has yet explored why, instead of which, students are suspended and whether suspensions may result from school characteristics as well as student behavior. Using stepwise regression, this study analyzed national data on students' suspensions, misbehavior, antisocial attitudes, perceptions of teacher interest, race, and Grade Point Average; teachers' race and perceptions of student problem-solving capacity and proportion of low-ability students; and the schools' proportion of white students, fairness and consistency in student governance, and degree of adherence to administrative rules or teacher discretion in handling discipline. Data came from questionnaire responses from 31,373 students, 641 principals, and 23,895 teachers in a stratified random sample of 641 public secondary schools drawn from the 1976 National Institute of Education (NIE) national Safe School Study. The results showed that suspension does have a basis in student attitudes and behavior but is more strongly affected by school factors. School characteristics were more likely to lead to suspensions where (1) teachers believed that students could not solve problems, (2) students perceived teachers as uninterested in them, (3) schools handled discipline largely by administrative rules, (4) school governance of students was inconsistent or unfair, and (5) school personnel had relatively high degrees of academic and racial bias. (Project Director/RW)