Researchers and policymakers increasingly recognize that the livestock sector supports the livelihoods of a large proportion of rural households in most African countries and may have an important role to play in rural poverty reduction strategies. In order to develop this insight, economywide models should capture both the biological, dynamic relationships between the stocks and flows of livestock and the economic linkages between the sector and the rest of the economy. We extend an existing dynamic recursive general equilibrium model for the Ethiopian economy which better models the livestock sector. A separate herd dynamics module enables us to specify stock-flow relationship, distinguishing between the capital role of livestock and the flow of livestock products. We also improve the underlying system of economic accounts, to better capture draft power and breeding stocks. We use this model to simulate separate, realistic Total Factor Productivity (TFP) shocks to three agricultural subsectors—cereals, cash crops, and livestock—and compare them with a baseline scenario replicating the 1998 to 2007 productivity trends, following Dorosh and Thurlow (2009) who have examined CAADP productivity scenarios. The results we obtain reveal the important role of the livestock sector in increasing various measures of GDP and combating food insecurity. Agricultural GDP and overall GDP growth levels achieved in the livestock TFP shock scenario are very similar to those achieved in the cereal TFP shock scenario, unlike what previously thought. Importantly, as factors are dynamically re-allocated between agricultural activities, our analysis highlights the inefficiency of strategies focusing on cereal sector development alone. Moreover, livestock sector productivity growth leads to greater factor income growth, particularly labor income, than in the other simulations. Labor is the predominant asset of poor households and hence large income gains and food consumption growth are realized under the livestock-led scenario.