The development of community-based services for offenders has not always been marked by adherence to best practice based on empirical evidence. Expediency, personal conviction and a sincere desire to do ‘something’ to respond to antisocial behaviour and criminality in our communities are historically more common drivers. This article explores some of the contemporary research and evidence relating to the efficacy of various approaches used in implementing offender rehabilitation programmes. The authors outline a particular model at work in the Cornmarket Project. Mindful of current economic constraints and the increased reporting requirements placed on organisations in receipt of state funds, the article emphasises the importance of having in place an easy-to use, valid and reliable outcome measurement system for programmes dealing with offender reintegration and rehabilitation.