The impossibility of current spatial database systems and GIS to handle spatial vagueness and imprecision has been recognized as an important problem in the spatial database domain. For years, researchers have focused on identifying the appropriate concepts that are necessary to effectively and efficiently deal with the vagueness and imprecision that not seldomly appears, but is widespread amongst spatial objects (especially those that are naturally occurring such as forests, mountains, and rivers). In this paper, we analyze the most popular formalisms for dealing with spatial uncertainty in databases. The analysis in this paper is centered around the definition of our Vague Spatial Algebra (VASA) which exemplifies a complete approach to handle vagueness in spatial databases. We compare VASA with existing concepts based on rough set theory and fuzzy set theory.