This paper examines the effects of private schooling on adolescent nonmarket behaviors. Differences between private- and public-school students were controlled by making use of the rich set of covariates available with the Nation Education Longitudinal Survey (NELS) microdataset. An instrumental-variables strategy was also employed that exploits variation across metropolitan areas in the costs that parents face in transporting their children to private schools, which stem from differences in the quality of the local transportation infrastructure. Evidence was found suggesting that religious private schooling reduces teen sexual activity, arrests, and use of hard drugs (cocaine), but not drinking, smoking, gang involvement, or marijuana use. Two questions require further research: Could public schools reduce teen sexual activity by either including "character education" as part of the curriculum without reference to specific religious ideologies or by dropping sex and drug education classes? and Are there other policies that public schools could adapt from religious private schools that would improve the academic or nonacademic outcomes of students? (Contains 55 references and 9 tables.) (RT)