Duty drawback schemes, which typically involve a combination of duty rebates and exemptions, are a feature of many countries' trade regimes. They are used in highly protected, developing economies as means of providing exporters with imported inputs at world prices, and thus increasing their competitiveness, while maintaining the protection on the rest of the economy. In China duty exemptions have been central to the process of trade reform and have led to a tremendous increase in processed exports utilizing imported materials. Despite the widespread use and importance of duty drawbacks, these "new trade liberalization" instruments have been given relatively little attention in empirical multilateral trade liberalization studies. This paper presents an empirical multi-region trade model GTAP-DD, an extension of GTAP, in which the effects of policy reform are differentiated based on the trade-orientation of the firms. Both GTAP and GTAP-DD are used to analyze the impact of China's WTO accession, which involves liberalization in China from 1997 to post-accession tariffs among a number of other liberalization measures. The analysis shows that failure to account of duty exemptions in the case of China's recent WTO accession will overstate the increase in : (a) China's trade flows by 40 percent, (b) China's welfare by 15 percent, and (c) exports of selected sectors by as much as 90 percent. The magnitude of the bias depends on the level of pre-intervention tariffs and the size of tariff cuts - the larger the initial distortions and tariff reductions, the larger the bias when duty drawbacks are ignored. The bias in GTAP's estimates of China's real GDP, trade flows and welfare changes due to WTO accession increases more three times when China's pre-intervention tariffs are raised from their 1997 levels to the much higher 1995 levels. These results suggest that trade liberalization studies focusing on economies in which protection is high, import concessions play an important role and planned tariff cuts are deep, must treat duty drawbacks explicitly in order to avoid serious errors in their estimates of sectoral, trade flows and welfare changes.