Targeted stigma and school voucher threats under a revised 2002 Florida accountability law have positive impacts on school performance as measured by the test score gains of their students. In contrast, stigma and public school choice threats under the US federal accountability law, No Child Left Behind, do not have similar effects in Florida. Estimation relies upon individual-level data and is based upon regression analyses that exploit discontinuities within the accountability regimes. Choice threats embedded within accountability regimes can moderate educational inequalities by boosting achievement at the lowest-performing schools, but policy design is crucial. Copyright 2006 Royal Economic Society.