Despite the popularity of policy briefs as a tool for disseminating research, there is no evidence of their effectiveness in changing people's beliefs. We conducted an experiment whereby readers of a policy brief were randomly assigned to different versions of the brief and to a control group. We collected data on opinions and knowledge regarding the impact of agricultural interventions before and after reading a brief disseminating the conclusions of a systematic review. We found that the brief helped some readers to form an opinion, but we found no evidence of a change in prior beliefs. We recommend that more trials and laboratory experiments should be conducted to assess the efficacy of policy briefs and attitudes to evidence-based policy-making. © 2013 Copyright Taylor and Francis Group, LLC.