Anyone caring to review recent literature on the use of community service as a direct alternative to custody might well conclude that it is failing as a tariff measure. In this article the author poses two questions which are directly relevant to this growing debate. Firstly, how valid is this conciusion in the light of differing sentencing policies amongst schemes? Secondly, where the policy is clearly to use community service as a tariff sentence, how can the number of offenders diverted from custody be adequately estimated?