Abstract: This paper reports the results of an experiment designed to assess the ability of an enforcement agency to detect and deter harmful short-term activities committed by groups of injurers. With ordered-leniency policies, early cooperators receive reduced sanctions. We replicate the strategic environment described by Landeo and Spier (2018). In theory, the optimal ordered-leniency policy depends on the refinement criterion applied in case of multiplicity of equilibria. Our findings are as follows. First, we provide empirical evidence of a "race-to-the-courthouse" effect of ordered leniency: Mild and Strong Leniency induce the injurers to self-report promptly. These findings suggest that the injurers' behaviors are aligned with the risk-dominance refinement. Second, Mild and Strong Leniency significantly increase the likelihood of detection of harmful activities. This fundamental finding is explained by the high self-reporting rates under ordered-leniency policies. Third, as a result of the increase in the detection rates, the averages fines are significantly higher under Mild and Strong Leniency. As expected when the risk-dominance refinement is applied, Mild Leniency exhibits the highest average fine.