The agricultural trade liberalization proposal known as 'tariffication' aims at converting all existing non-tariff barriers (NTBs) to trade into bound tariffs, and to reduce these tariffs over time. This is in tune with the original philosophy of the General Agreement on Tariffs and Trade (GATT) and it calls for a dramatic overhaul of existing agricultural policies in many developed countries. The main economic issues that arise with tariffication stem from the non-equivalence of tariffs and NTBs in a number of scenarios. This paper analyzes non-equivalence arising from the existence of: imperfect competition in importing countries; price instability in importing and exporting countries; and, inefficient allocation of quantitative restrictions. It is shown that in all these cases the definition of an appropriate 'equivalent tariff to be used in tariffication is not straightforward, and that in general this equivalent tariff cannot be computed on the basis of only observed price differences between countries. Tariff-rate quotas, which are meant to be the main tool of implementation of tariffication according to the existing proposal, are analyzed in some detail. Concerning the relationship between tariffication and the other elements of the trade liberalization package, it is shown that tariffication would limit the scope of export subsidy policies. It is also shown that the existence of production and export subsidies makes observed price gaps between countries of questionable value in setting equivalent tariff levels. Finally, it is argued that the main focus of tariffication should be the conversion of NTBs to acceptable long-run (bound) tariffs rates, and considerable flexibility in this conversion process could be exercised in the transition period.