It is well-known that using performance measurement systems in public services can lead to dysfunctional consequences even when people operate with the best of intentions. Nevertheless, the same mistakes are made time and time again. In the hope that understanding why things go wrong may help reduce future mistakes, this paper examines some of the underlying reasons for this dysfunctionality, applying a range of theoretical insights to do so. It bases its argument on UK experiences, but these are transferable to other countries intent on similar measurement schemes. Its argument will be of value to those who wish to improve the performance of performance measurement systems in public services, to those who wish to understand why things can go wrong and to those who'd rather not be hurt when things go wrong.