The remit of this paper is to investigate the role of prefixes of spatiality in the formation of words. To do so, it adopts two approaches. Theoretically, it adopts Cognitive Semantics (CS) and tries to substantiate some of its tenets with reference to prefixation. One tenet is that linguistic items are meaningful. On this basis, it argues that prefixes of spatiality have a wide range of meanings that gather around a central sense. Another tenet is that the meaning of a linguistic item is best understood in terms of the domain in which it is embedded. On this basis, it argues that prefixes of spatiality form a set which highlights not only similarities but also differences. A further tenet is that the use of an expression is governed by the particular construal imposed on its content. On this basis, it argues that a derived word is used relative to the particular construal the speaker chooses to describe a situation. Empirically, it adopts Usage-based Linguistics, and tries to substantiate some of its prmises with reference to prefixation. One premise is that knowledge of language is derived from language use. Another premise is that the linguistic system is shaped by actual data. A further premise is that structure and substance are tightly linked.