Possible worlds and propositions are the most fundamental building blocks of intensional semantics, as well as the most fundamental building blocks of modal logic. Within the standard possible-world semantics there are two prevailing approaches to the explication of possible worlds and propositions. The first approach treats possible worlds as primitive and specifies propositions in terms of them (propositions as sets of primitive possible worlds in which they hold). The second approach treats propositions as primitive and specifies possible worlds in terms of them (possible worlds as maximal consistent sets of primitive propositions). Supposing we wish to stay within the standard possible-world setting, the aim of this paper will be to compare these two approaches: Which arguments have been (can be) listed in their favour? Can these arguments help us to decide between them? It should be clear that the present paper is not going to explicate any modal notions (such as necessity, obligation, belief, and so on); its perspective will not be semantic or logical, but rather methodological.