Until 2050, the global population is projected to reach almost 9 billion people resulting in a rising demand and competition for biomass used as food, feed, raw material and bio-energy, while land and water resources are limited. Moreover, agricultural production will be constrained by the need to mitigate dangerous climate change. The agricultural sector is a major emitter of anthropogenic greenhouse gases (GHG). It is responsible for about 47 % and 58 % of total anthropogenic emissions of methane (CH4) and nitrous oxide (N2O) (IPPC, 2007). CH4 emissions are associated with enteric fermentation of ruminants, rice cultivation and manure storage; N2O emissions are related to nitrogen fertilizers and manure application to soils, but also to manure storage. Land use changes, pasture degradation and deforestation are the main sources of agricultural CO2 emissions, where livestock is a major driver of deforestation and climate change, accounting for 18 % of anthropogenic GHG emissions (Steinfeld et al., 2006). In this context, the key role of livestock is to be investigated. According to FAO, livestock uses already about 30% of the Earth‘s land surface as resource for grazing while demand for livestock products will continue to rise significantly, especially to feed the animals. For the assessment of future food supply and land-use patterns as well as the environmental impacts of the agricultural sector, there is an urgent need to identify and analyse main characteristics of the livestock sector. Concerning the conversion efficiency of natural resources like land and water to animal products, feeding technologies play a crucial role. They also determine the magnitude of environmental impacts per amount commodity generated. For ten world regions we define the feeding technology for five livestock subsectors as a set of the following parameters: feed mix, feed energy requirements per unit output, and methane emissions per unit output. We calculate these parameters on the basis of FAO Food Balance Sheets and data from the literature. The resulting regional feed demand of marketable feed is consistent with FAO data. To assess the impacts of different feeding technologies, we implement this concept in the global land use model MAgPIE that is appropriate to assess future anthropogenic GHG emissions from various agricultural activities and environmental and economic impacts of different pathways of the agricultural sector by combining socio-economic regional information with spatially explicit environmental data. We compare three alternative feeding scenarios in terms of GHG emissions from agricultural activities (CH4, N2O). We find that methane emissions rise significantly under a scenario of production extensification (i.e. higher roughage shares in feed mixes). Under an intensification scenario, future methane emissions are even lower than in 1995, but N2O emissions from nitrogen fertilizers and manure application to soils increase.