Skeptics of the KIPP (Knowledge Is Power Program) charter school network argue that these schools rely on selective admission, attrition, and replacement of students to produce positive achievement results. We investigate this using data covering 19 KIPP middle schools. On average, KIPP schools admit students disadvantaged in ways similar to other local students, and attrition patterns are typically no different at KIPP than at nearby schools. Unlike district schools, however, KIPP schools tend to replace students who exit with higher achieving students, and fewer students are replaced in the later years of middle school. Overall, KIPP's positive achievement impacts do not appear to be explained by advantages in the prior achievement of KIPP students, even when attrition and replacement patterns are taken into account.