An increasing share of new teachers participate in alternative preparation programs. A fundamental question is what effect these programs have on long-run student achievement. To address this question, I study Teach For America (TFA) teachers working in New York City (NYC). I first estimate within-teacher returns to experience and find that TFA alumni teachers improve at double the rate of the average NYC teacher over their first 5 years in the classroom. I then model the joint relationship between turnover and performance over time. The results suggest that the net effect of TFA hiring on long-run student achievement is positive, despite the regular churn within the TFA workforce.