This study focuses on the link between agricultural trade openness and the sector's performance, an improvement in which could have significant impacts on poverty reduction. We emphasize Latin America, during the 1960-2005, using a recently constructed data base of agricultural support Nominal and Relative Rates of Assistance (NRA and RRA) that includes information for several developing countries, beyond the region. The principal question addressed is, does the trade regime influence sectoral growth? With the answer to this question we then make some inferences regarding the influence of sectoral growth on poverty, using estimates of the impact of agricultural growth on national economic growth which in turn impacts the incomes of the poorest quintile. The empirical analysis takes advantage of cross-country panel data from several sources, covering many developing countries in Africa, Asia and the LAC region. The LAC countries are Argentina, Brazil, Chile, Colombia, Dominican Republic, Ecuador, Mexico, and Nicaragua. We compare groups of countries, defined by their levels of protection and changes in those levels (using both NRA and RRA), to assess the effects of the trade regime on growth in agricultural value added and production (using FAO's production index). A panel data regression analysis is also presented to estimate the impacts of the levels and changes in protection. The findings are: First, based on both the comparison of country groups and the regression analysis, when explaining agricultural GDP or production indices changes in the trade regime are more important than the absolute values of the protection levels themselves. Second, based on the regression analysis, for a representative country, removing the taxation of the trade regime (prevailing in the 1970s and 1980s) would have resulted in an increase over trend growth (at least over a five-year horizon) in the average agricultural GDP growth of about 50%. Third, using the regression model estimates and previous estimates of the links between agricultural growth and national growth and income of the poorest quintile, we simulate what would have been the impact on the income of the poorest if a representative high-tax country (a negative and stable NRA) had moved to a neutral trade regime (an NRA of zero). Annual average income growth would have risen approximately one-quarter point, or about 9 percent over its average rate during the subsequent five-year period. We offer arguments why this is likely a low estimate. Finally we discuss the implications for a future policy agenda, especially in light of the large number of LAC countries which still have high levels of interventions, both positive for importables and negative for exportables, although average sectoral protection indicators are now relatively small.