It is often argued that institutionalized after-school care (ASC) can benefit children lacking adequate homework support at home and, hence, foster equality of opportunity. However, despite considerable policy interest, it is unclear whether these afternoon programs are beneficial for child development and if selection into them is efficient, i.e., whether students benefiting most from the programs choose to attend. In this paper, I examine the effects of ASC on elementary school children’s schooling outcomes and non-cognitive skill development. Using a marginal treatment effect framework and regional and temporal variation caused by an extensive reform in Germany, I instrument after-school care attendance with the change in the distance to the next school offering ASC within one district. My findings suggest that children from lower socioeconomic backgrounds, who more often select into treatment, have higher ASC premiums. Treatment effects on the treated’s non-cognitive skills are more sizable than those on the untreated, suggesting that selection into ASC is positive and efficient. Overall, a universal voluntary offer of ASC will likely help reduce educational inequalities.