This paper deals with an experiment by the Swedish Tax Agency to test the effect of information to taxpayers regarding different audit strategies . The experiment involved approximately 900 sole proprietors, divided into three groups, where one was informed that audits would focus on taxpayers declaring the lowest income, i.e. according to a rational audit strategy. Another group was told that audits would be made at random whereas the third was a control group. The effect of strategy information was measured as the change in declared income between years. The principal finding was that declared income increased significantly more in the rational-audit-strategy group than in the control group.