The spaghetti bowl phenomenon expected from the proliferating East Asian regional trade agreements (RTAs) is worrisome. In particular, the complicated web of hub-and-spoke type of overlapping free trade agreements (FTAs) can result in high costs for verifying rules of origin. As an alternative policy option to avoid the negative effect of trade deflection, customs unions (CUs) should be examined. Most of the theoretical analyses on the formation of CUs highlight stronger positive welfare effects compared to FTAs. However, there is a lack of empirical evidence to support the second-best theory of customs unions. This paper is an attempt to fill this gap by applying two methodologies: an ex ante simulation approach and an ex-post econometric approach. We quantitatively estimate the trade effect of CUs and FTAs by adopting a Gravity regression analysis. In general, we find that a CU is a superior type of RTA to an FTA in terms of creating more intra-bloc trade. In addition to analyzing the trade effects of RTAs according to type, we quantitatively evaluate the welfare and output effects of CUs for East Asia (an ASEAN+3 CU and a China-Japan-Korea CU) compared to FTAs by applying a computable general equilibrium model analysis. The East Asian CUs adopt a system of common external tariffs (CET) based on simple-averaged, import-weighted, consumption-weighted, and minimum rates. Overall, we find that the ASEAN+3 CU with the minimum CET are the most desirable type of RTA for both East Asian member countries and the world economy as a whole.