Criminals often engage in costly avoidance to lower their likelihood of being detected and sanctioned. Such avoidance, in turn, affects the optimal enforcement policy. This paper studies how the optimal standard of proof is affected by a specific type of avoidance, namely investments by guilty defendants to mimic innocent individuals. We show that when mimicking is possible the optimal standard of proof is weaker than preponderance of the evidence. This stands in contrast to much of the literature, which shows that non-deterrence costs usually cause the standard of proof to be stronger than preponderance. Thus, it is important not to ignore criminals’ secondary behavior when determining the optimal standard of proof.