Using data from the National Education Longitudinal Study, this report examines eighth grade school variables that might predict college attendance for students whose parents have no college education (n = 2,521). The selected school variables are amenable to public policy and/or skilled leadership, and they include "school structure" (e.g., number of school days in an academic year), "student statistics" (e.g., attendance rate), "atmosphere" (e.g., seriousness of student problems), "academic offerings" (e.g., required year-long courses for core subjects), and "teacher statistics" (e.g., level of absenteeism). Controlling for students' sociodemographic characteristics and past academic achievement, multivariate logistic regressions show that higher teacher absenteeism in eighth grade lowers the odds of attending college for potential first-generation students. (Contains 4 tables.)