A procedure is developed to examine the ex-post impacts of improved maize varieties on poverty in rural Ethiopia. Yield and cost effects of adoption are estimated econometrically under assumptions of both homogeneous and heterogeneous treatment effects. A backward derivation procedure is employed within an economic surplus framework using estimated treatment effects to identify the counterfactual income distribution without improved maize arieties. Poverty impacts are estimated as the differences in poverty indices computed using observed and counterfactual income distributions. Improved maize varieties have led to noticeable reduction in the poverty headcount ratio, depth, and severity in rural Ethiopia. However, poor producers benefit the least from adoption because their land areas are limited.