A study published in the current issue of Health Affairs on Medicare’s flagship hospital pay-for-performance program casts doubt about whether pay-for-performance strategies that reward improvement can generate greater improvement among lower performing providers. The program – Premier Hospital Quality Incentive Demonstration – began in 2003 but changed its incentive design in late 2006. The goals were to encourage greater quality improvement, particularly among lower-performing hospitals. However, the authors found no evidence that the change achieved these goals. Although the program changes were intended to provide strong incentives for improvement to the lowest-performing hospitals, they found that in practice the new incentive design resulted in the strongest incentives for hospitals that had already achieved quality performance ratings just above the median for the entire group of participating hospitals. Yet during the course of the program, these hospitals improved no more than others. Findings from this study raise questions about whether pay-for-performance strategies that reward improvement can generate greater improvement among lower performing providers. They also cast some doubt on the extent to which hospitals respond to the specific structure of economic incentives in pay-for-performance programs.