In this paper we develop a dynamic CGE model to examine the impact of CAFTA on production, employment and poverty in El Salvador. We model four aspects of the agreement: tariff reductions, quotas, changes in the rules of origin for maquila and more generous treatment of foreign investment. The model shows that CAFTA has a small positive effect on growth, employment and poverty. Tariff reduction under CAFTA adds about .2% to the growth rate of output up to 2020. Liberalizing the rules of origin for maquila has a bigger positive effect on growth and poverty mainly because it raises the demand for exportables produced by unskilled labor. We model the foreign investment effect by assuming that capital inflows go directly to capital formation. This raises the growth rate of output by over 1% per year and lowers poverty incidence in 2020 by over 25% relative to what it would be in the baseline scenario. These simulations say something important about the growth process in a country like El Salvador in which it seems reasonable to assume that there is idle unskilled labor willing and able to work at a fixed real wage. In such an economy, growth can be increased in one of three ways. First, already employed resources can be moved to sectors where they are more productive. That is what the tariff reductions under CAFTA do, and the result is positive but small. Second, the structure of demand can be changed in such a way as to increase the demand for previously unemployed unskilled labor. That is what the maquila simulation does, because maquila uses a lot of unskilled labor relative to skilled labor and capital. Finally the supply of capital can be increased by increasing the rate of capital formation. That is what happens in the FDI simulation.