This paper shows that research on international agricultural trade reform can make much greater contributions to understanding than was feasible in earlier trade negotiations. Part of this is due to improvements in the basic data on production, consumption and trade associated with the development of the GTAP database and part due to the greater availability of disaggregated data on applied and bound rates of protection. Unfortunately, the framework in which we operate means that we are reduced to measuring gains that are substantially less than one percent of GDP. New developments in theory and method provide the potential for our quantitative analysis to be improved in at least six areas, including improvements in: the measurement of protection; incorporation of barriers to trade in services; representation of the counterfactual; aggregation of barriers; incorporating the emergence of new products; and reflecting the productivity enhancement associated with trade reform. Urgent action to improve our measures is needed to avoid our measures becoming either irrelevant or counter-productive in the process of policy formulation.