The present Conservative administration has attempted to improve economy, efficiency and effectiveness in government by decentralising the delivery of public services. In order to overcome the familiar organisational paradox, that to decentralise it is necessary to centralise, the government has encouraged the development of performance indicators as a ‘hands off’ instrument of control over service delivery. This article analyses various problems raised by the use of performance indicators as a tool of control. It is argued that performance indicators remain an imperfect mechanism of central control and still involve a much resented degree of ‘backseat driving’ by government.