Charter school effects remain uncertain. Small lottery studies on high-performing charters produce impressive results, but large observational studies on the full range of charter schools are less encouraging. To make matters worse, these observational studies that aim for representativeness are based on only switchers, a small and unrepresentative population, arguably defeating the purpose of this design. In the first study of charter school effects on the full range of charter school students in North Carolina, we find that, in general, charter school performance has improved over time, although often it continues to remain lower than traditional public school achievement. However, we find some evidence suggesting that black and economically disadvantaged students experience slightly more achievement growth in charter schools, particularly in reading, than do students who are white and not economically disadvantaged.