Includes bibliographical references.

We analyze all but a few of the 47 charter schools operating in New York City in 2005-06. The schools tend locate in disadvantaged neighborhoods and serve students who are substantially poorer than the average public school student in New York City. The schools also attract black applicants to an unusual degree, not only relative to New York City but also relative to the traditional public schools from which they draw. The vast majority of applicants are admitted in lotteries that the schools hold when oversubscribed, and the vast majority of the lotteries are balanced. By balanced, we mean that we cannot reject the hypothesis that there are no differences in the observable characteristics of lotteried-in and lotteried-out students. Using the lotteries to form an intention-to-treat variable, we instrument for actual enrollment and compute the charter schools' average treatment-on-the-treated effects on achievement. These are 0.09 standard deviations per year of treatment in math and 0.04 standard deviations per year in reading. We estimate correlations between charter schools' policies and their effects on achievement. The policy with the most notable and robust association is a long school year--as long as 220 days in the charter schools.