Recent theoretical developments show the conditions under which it is cost-effective for the regulator to induce perfect compliance in cap-and-trade programs. These conditions are based on the ability that a regulator with perfect information has to induce the firms to emit any desired level with different combinations of the number of permits supplied to the market and the monitoring probability, assuming that firms are expected profit maximizers. In this paper, we test this hypothesis with a series of laboratory experiments. Our results suggest that firms may behave significantly different from what these models predict precisely when the different combinations of the supply of permits and the monitoring probability induce compliance versus noncompliance. More specifically, by allowing noncompliance in a manner consistent with theory, the regulator could produce a decrease in emissions and an increase in the market price of tradable permits that is not predicted by the theoretical models. The implications for the cost-effective design of environmental policy are discussed.