Abstract: We consider a principal-agent model in which the principal can monitor and punish the agent with a fine if the agent is caught being untruthful. To reduce the probability of being verified, the agent can engage in costly avoidance. We design the optimal regulatory policies with and without avoidance. The optimal mechanism with enforcement allocates the object more often than the optimal mechanism without enforcement. Moreover, enforcement increases the expected transfers to the principal. Avoidance has two implications to the optimal regulatory mechanism: (i) the expected optimal transfers to the principal decrease and (ii) the principal allocates the object to a smaller share of types. If the latter effect dominates the former, it is possible that the agent's capability to engage in avoidance is disadvantageous not only for the principal, but also for the agent ex ante.