Technological change in agriculture plays a decisive role for meeting future demands for agricultural goods. Especially in the longer run, i.e. several decades, technological change will be one of the major determinants of agricultural production. However, up to now, most agricultural sector models and models on land use change have used technological change as an exogenous input due to various information and data deficiencies. This paper provides a first attempt towards an endogenous implementation based on a measure of agricultural land-use intensity, called τ -factor. We relate this measure to empirical data on investments in technological change as well as production costs. Our estimated yield elasticity with respect to research investments is 0.24 and production costs per area increase linearly with an increasing yield level. Having implemented this approach in the global land-use model MAgPIE (”Model of Agricultural Production and its Impact on the Environment”) we are able to project about required yield growth rates in the future. Highest future yield increases are obtained for Sub-Saharan Africa, the Middle East and South Asia. A validation with FAO data for the period 1995-2005 shows that our model behaviour is in line with recent observations.