The idea of reducing public punitiveness through providing information and encouraging deliberation has attracted considerable interest. However, there remains no solid evidence of durable changes in attitude. The study presented here provides a test of the hypothesis that information combined with deliberation can affect general measures of punitiveness, confidence in the courts and acceptance of alternatives to imprisonment (the three dependent variables). The study involved a pre-test, post-test experimental design. Participants were randomly allocated to either an intervention group or a control condition. Statistically significant changes in the dependent variables were observed immediately following the intervention but these changes were not sustained when measured at follow-up nine months later. Further, at the time of the follow-up the differences between the control group scores and the intervention group scores were not significantly different. The observed changes immediately following the intervention are seen to be a function of the changed relationship of the respondent to the task. The implications of the results for integrating public perspectives into policy are discussed. It is argued that rather than a focus on public education, a more productive direction is to focus on the way the public is engaged on matters concerning punishment.