NAFTA Rules of Origin (ROO) are used to determine which goods are attributable to NAFTA member countries and thus eligible for a preferential tariff. These rules create incentives for NAFTA producers to source inputs from potentially higher priced North American suppliers. As such, a ROO is an implicit tax on the intermediate goods produced by the rest of the world. Most computable general equilibrium (CGE) studies assessing the welfare impact of moving from NAFTA to a deeper form of integration, for example a North American Customs Union (CU), typically proxy the integration as the adoption of a common external tariff towards the rest of the world. Thus, these studies do not explicitly consider the impact of eliminating the distortion created by the NAFTA ROO. This paper shows that the failure to account for the removal of ROO in these studies would likely lead to biased estimates. I explicitly consider the impact of removing the NAFTA ROO in a multi-country multi-sector dynamic general equilibrium model. The objective is to provide better estimates of this shock on production and welfare. Although the removal of distortionary ROO is likely to lower the unit cost of production within NAFTA, it may also deteriorate NAFTA terms of trade with the rest of the world. The net effect on welfare is ambiguous and is thus an empirical issue, which is addressed in this paper. This is illustrated using three distinct scenarios for which I do not take sides, but that I propose as a springboard for a general discussion.