I study an expansion of a South African social grant program that provided unconditional cash transfers to adolescents for the first time. Over the period 2009 to 2012, age eligibility for the child support grant was progressively extended from children under 14 to children under 18 years old. I use a difference-in-difference identification strategy that exploits the cross birth cohort variation in adolescent grant eligibility generated by these age eligibility changes to examine how unconditional cash transfers affect schooling outcomes in adolescence and young adulthood. I find that adolescent grant eligibility increases enrollment and attainment, with the effects concentrated among females, rural individuals, and those with the lowest numerical literacy. I explore education spending as a channel through which the child support grant affects education outcomes.