Uses a longitudinal cohort of eighth graders to examine whether measures of school engagement predicts school-age pregnancy among white, black, and Hispanic teens.