This report presents a research project on 'evaluated road safety campaigns', funded by the European Commission under the Transport RDT Programme of the EU's Fourth Framework Programme. The project's main objectives were to: (1) collect a large international sample of campaign evaluations; (2) accurately describe this sample, using a detailed coding scheme and focusing on the design of the evaluation; (3) assess some effects of the campaigns on reducing accidents; and (4) suggest policy guidelines for future road safety campaigns. A pragmatic strategy was developed, aimed at developing a reliable description of evaluated road safety campaigns. The third stage used a meta-analysis, a quantitative method of combining findings across studies of the same subject, even where the studies used different measures to assess the same dimension. The meta-analysis was based on 35 studies with 72 results, 52 of which were for before-during studies and 20 of which were for before-and-after studies. It was estimated that the average reductions in accident numbers were 8.5% during the campaign period, and 14.8% after the campaigns; both these estimates are statistically significant. These results must be attributed to all components of the campaigns, not only their media and publicity aspects. Several recommendations are made.