This dissertation investigates the role of geophysical, agroecological, and socioeconomic determinants of land-use change during the last 25 years in two districts of Dak Lak province in the Central Highlands of Vietnam. The analysis of these determinants allows to assess the influences of various rural development policies on land-cover changes. Landsat satellite images from the same cropping period of the years 1975, 1992 and 2000 are interpreted to detect land-cover change between the two time periods. A survey in randomly selected villages provides primary recall data on socioeconomic and policy variables hypothesized to influence land-use change. Secondary data on rainfall, soil suitability, and topography were obtained from meteorological stations, from a digital soil map and a digital elevation model. All data were spatially referenced using GIS software. Survey data is merged with spatially explicit raster data using accessibility catchments, which are designed to approximate village areas based on the estimated travel costs from each cell to the village location. A reduced-form, multinomial logit model is used to estimate the influence of hypothesized determinants on land use and the probabilities that a certain pixel has one of five land-cover classes during either of the two periods under consideration.Results suggest that the first period from 1975 to 1992 was characterized by land-intensive agricultural expansion and the conversion of forest into grass and agricultural land. During the second period, since 1992, the rapid, more labor-and capital-intensive growth in the agricultural sector was enabled by the introducti-III IV ABSTRACT on of fertilizer, improved access to rural roads and markets, and the expansion of irrigated areas. These policies, combined with the introduction of protected forest areas and policies discouraging shifting cultivation during the second period reduced the pressure on forests while at the same time increasing agricultural productivity and incomes for a growing population. Forest cover during the second period mainly increased due to the regeneration of areas formerly used for shifting cultivation.Policy simulations mimic the influences from potential policy interventions on land use. Land-cover categories for the policy simulations are aggregated to three classes and a spatial sample is drawn to concentrate the analysis on changes influenced by anthropogenic interventions. In that way, the focus of the simulations is on land-use changes within already cultivated areas and at the agricultural frontier where most land-use changes relevant for policy makers take place. The simulations are limited to the second period from 1992 to 2000 as this is the relevant period for potential policy measures.The four rural policy scenarios are carried out for low and high levels of investments in irrigation facilities, for an enlarged area under forest protection and for a combination of high investments in irrigation combined with increased forest protection, an often mentioned...