This paper explores the prospect of coupling fuzzy set theory and the analytic hierarchy process (AHP) into a raster-based geographical information system (GIS) for the preliminary screening of landfill sites in Thailand. The theory of linguistic variable is used to represent imprecision of spatial data and human cognition over the criteria used for the screening process. Proximity of geographic objects, slope and elevation are criteria used for this investigation. The priority weights reflecting the preferences on the screening criteria, accounting for seventeen map layers, are derived by the method of the AHP. The general method of GIS intersection based on binary logic is conducted to compare with the fuzzy min-operator intersection and the proposed convex combination model. The results show that the first method, in relation to the fuzzy methods, fails to recognize 35.6% of the study area as potential areas for waste disposal. The proposed convex combination model has advantage over the fuzzy min-operator intersection with respect to the ability to integrate criteria's preferences into the screening process. It also yields agreeable results with the recommendations from previous study in the same area.