In Malawi, the pregnant schoolgirl embodies failure for diverse actors and institutions. She signals moral degeneration and a loss of control over girls' sexuality for parents and teachers, chiefs and clerics. At the same time, she demonstrates the programmatic failure of schooling to delay reproduction and trigger a "ripple effect" of positive social, demographic, and economic outcomes touted in mainstream development discourse (Moeller, 2014). My dissertation explores these tensions by focusing on Malawi's 1993 Readmission Policy, which banned the practice of permanently expelling pregnant girls from school, its failure, and its 2016 reform. Readmission Policy serves as a node through which to examine why, though young mothers in Malawi have been allowed to re-enroll in school after delivery for over two decades, very few do. And it offers a timely opportunity to interrogate longstanding frictions around production, reproduction, and the constitution of power in the gendered project of "development." Using multi-sited ethnographic methods and an anthropology of policy approach (Shore, Wright, and Pero, 2011), my research uncovers how diverse actors from pregnant students to school personnel and staff in Malawian nonprofits, ministry offices, and international funding agencies conceptualize gender and the relationship between fertility, sexuality, and schooling. I theorize the pregnant schoolgirl as a "boundary subject" and explore the overlapping meanings assigned to her. I trace how "she" shapes what stakeholders say, write, and do as they reform Readmission Policy and craft an approach to gender in schools. With a focus on processes of development, I move beyond studies of whether development works to understand how it works (Mosse, 2004, 640), highlighting throughout the constitution and circulation of power among unequally situated actors and institutions.