We exploit information about parental preference and a randomized component in the assignment of students to schools within a deferred acceptance (DA) mechanism to estimate the causal effect of enrolling in a charter school in Newark, New Jersey, on student test scores. The estimates incorporate variation from students attending about 70% of the city's charter schools, accounting for about 85% of charter school enrollment. Enrolling in a Newark charter school that participated in the DA assignment process leads to a large and statistically significant increase in math and English Language Arts (ELA) scores. Enrolling in a charter school that is operated by either the Knowledge Is Power Program (KIPP) or Uncommon charter school networks has an especially large effect.