Public agencies in the UK and elsewhere are increasingly required to set outcome targets as a strategy for improving their services. A crucial element of this ‘results orientation’ is a clear definition of the desired outcomes and a specification of appropriate performance indicators. A recent example of this policy in the UK-Local Public Service Agreements (LP S As)-is examined in this article. The authors’ analysis of the first generation of LP S As shows that just under half of the indicators used were measures of outcome. The authors explain the ‘wicked’ issues in outcome measurement that emerged from the research.