Over the last 10 years the use of Special Trade Preferences – the granting of lower import tariffs to specific countries – has increasingly been used as a tool to encourage development in developing countries, primarily in Africa. Yet, despite the popularity of these tools - practically all sub-Saharan countries receive some form of special trade preferences - we have little understanding of how well they work, both in terms of transferring wealth and encouraging growth. On what basis should they be evaluated? How should they be designed to be the most effective? Can they be superior to alternatives of traditional foreign aid?We seek to answer these questions in three ways. First, we develop a theoretical model that specifically allows for varying types of export goods, in particular in the use of labor and (potentially imported) inputs as requirements on the use of imported intermediated inputs is an important feature of special trade preferences. Second we use two changes in trade policy – the removal of tariffs on fabric for a number of African Countries with the African Growth and Opportunity Act (AGOA) of 2000 as well as the expiration of fabric quotas primarily for Asian competitors in 2004 – to test the predictions of the model. To test the data we will use a highly disaggregated data set containing information on volume, type of imports and price, of imports into the US from a large set of exporters. Third we design a theoretical model that allows for growth enhancing effects of policy and study optimal policy in this context.Preliminary theoretical results indicate that the current design of AGOA – with no requirements on third party inputs – encourages the production of products with a large imported share of fabric and the removal on quotas for Asian competitors in 2004 pushed African exports into cheaper exports. Both of these results find preliminary support in the data.