The WTO’s Doha Development Agenda has generated demand for estimates of the potential economic consequences of global trade reform. Recent improvements in the GTAP dataset have provided a much better representation of tariff restrictions as of 2001. However, despite its use by most global trade modelers, substantial differences in results emerge from different computable general equilibrium exercises. To help understand these differences, this paper examines the sensitivity of full global and regional trade liberalization results from the GTAP model, using the GTAP version 6.1 database, to different assumptions about factor mobility, fiscal neutrality, macro-economic closure, and trade (Armington) elasticities.