The U.K. National Audit Office (NAO) commissioned RAND Europe to conduct this review to identify and synthesize international research about the effectiveness of community orders in reducing re-offending. In this report, we review research on ten of the common requirements contained in community orders. Through examining reviews, systematic reviews and meta-analyses we draw conclusions about the state of research in the areas of unpaid work, mental health treatment, education/skills training, drug treatment, anger management, alcohol treatment, programmes for perpetrators of domestic abuse, regular probation, intensive probation and cognitive/behavioural programming. We also assess the strength of the evidence on whether each of these requirements affects the likelihood of re-offending. We find that the quality of research on the effectiveness of community-based interventions for offenders is extremely variable. However, in two areas (cognitive/behavioural programming and drug treatment) rigorous research exists which points to a reduction in the odds of re-offending. In four other areas (programmes for domestic abuse perpetrators, unpaid work, education and basic skills training and intensive probation) existing studies have not suggested that the programmes have a positive effect on recidivism. Finally, in four areas (anger management, probation, and alcohol and mental health treatment) the question of impact on re-offending remains unsettled. This review highlights the need for more rigorous research, especially randomized trials, into the requirements that constitute community orders. [For associated report, see ED502727.]