This study used the nation-wide U.S. Department of Education’s Office of Civil Rights dataset to investigate the school characteristics that influence disproportionate rates of harsh discipline in public schools. At the individual student level, research indicates that racial minority students receive suspensions and expulsions at rates that are greater than their membership in their schools’ population. However, there is little research investigating whether school-level variables (i.e., school characteristics) can predict disproportional discipline and how predictors may differ between groups. This study used school-level discipline data from schools in all fifty states as well as the District of Columbia with a minimum enrollment of 15 each of African American, Hispanic and White students to investigate potential school characteristics that predict disproportional discipline as well as whether the effect sizes of these predictors vary for disproportionality with regard to African American and Hispanic students. Predictors were school-level and included student body size, diversity and poverty level; school level and typology; student: teacher ratio and percentage of new teachers, and the presence of school resource officers, while states were used as a clustering variable and controlled for state-level effects such as state discipline policies. Results indicated that on average, disproportional discipline continues to affect both African American and Hispanic students, but that the degree of disproportionality varied considerably between schools. Few of the included variables predicted disparities in African American discipline. However, several variables, including the percentage of minority student enrollment, percentage of Hispanic student enrollment and being a middle school, high school or rural school, predicted disparities in Hispanic discipline. Results are interpreted and discussed in the context of existing, relevant research literature.