Global land use change is influenced by changes in diets towards more livestock products as well as additional demand for biomass from a growing biofuels industry. Global simulation models trying to quantify future food demand and subsequent land use change often differ in their results due to different demand specifications and elasticities. In order to isolate the impact of the latter on food demand and subsequent land use change we calibrate the global CGE model DART-BIO to different demand specifications and test the effect of changing income elasticities for livestock products. Our findings confirm the importance of demand specifications and income elasticities in influencing final models results, as the size of fixed subsistence quantities in the linear expenditure system (LES) determines the overall flexibility of demand. Large subsistence quantities dampen the responsiveness of private demand to income and price changes and lead to unrealistic reductions in global land use. As income elasticities of livestock demand in China and India become less elastic with growing income, the subsequent decrease in demand for livestock products and fodder crops affects land use worldwide. In general, there is some land use change at the expense of either soybeans or wheat in favor of oilseeds in all regions although the percentages remain very small. In addition, our results imply that the positive impact of biofuel production on livestock output as found by other studies is not reciprocal.