Unexpected variation in emissions can have a substantial impact on the prices and efficiency oftradable emission permit markets. In this paper we report results from a laboratory experiment inwhich subjects participate in an emissions trading market in the presence of emissionsuncertainty. Subjects face exogenous, random positive or negative shocks to their emission levelsafter they make production and emission control plans. In some sessions we allow subjects tobank their unused permits for future use. In all sessions, subjects can trade in a reconciliationperiod to buy or sell extra permits following the shock realization. Subjects then report theiremissions to the regulatory authority and they are placed in different inspection groupsdepending on their compliance history. The design of our experiment allows us to identifyimportant interactions between emission shocks, banking, compliance and enforcement. We findthat the relationship between emission shocks and price changes is significantly stronger withoutbanking, so banking helps smooth out the price variability arising from the imperfect control ofemissions. This greater price stability comes at a cost, however, since noncompliance andemissions are significantly greater when banking is allowed.