Abstract. The need for location-based services has motivated an important research effort in the efficient processing of location-dependent queries. Most of the existing approaches only deal with locations at maximum precision (e.g., GPS coordinates). However, due to imprecision or expressivity requirements, there are situations in which locations must be handled at different granularity levels (e.g., neighborhoods, cities, states, etc.). Indeed, whenever a set of locations are represented together as a granule, a meaning is implicitly given to the set. So, the use of different granularities brings different semantics to the location data. In this paper, we propose the use of semantic location granules to enhance the expressivity of location-dependent queries. This is done by exploiting the semantic information that is asserted about different granularity levels. This information could be, for example, the cost incurred by a moving object to traverse a spatial area or a requirement to ...