The effects of trade liberalization on poverty in Tunisia are examined, using a layered dynamic CGE-microsimulation approach. A dynamic CGE model endogenously generates the evolution of prices and, for each household group, income paths under protection and freer trade assumptions. These results are then used to assess the equivalent income of each household, using a sample from 1995 household survey, and so the effects of the simulated changes on poverty. Dominance tests are also used to avoid the arbitrariness of choosing a poverty line and a poverty measure. Simulation results show that although trade openness slowdowns the downward trend of poverty in the short and medium-run, it enhances poverty reduction in the long-run.