This thesis is devoted to an investigation of Aristotle’s notion of homonymy. When first introducing this notion, Aristotle tells us that things are homonymous when they have their name in common but the definitions corresponding to the name are different. (Categories 1a1–2) Homonyms are thus opposed to synonymous things which are things that have both name and the definition corresponding to the name in common. Clearly, then, river banks and money banks are homonymously called banks. But what about healthy things? Are bananas, lifestyles and organisms synonymously or homonymously called healthy, or maybe neither? Can it be that they fall into a category occupying the middle ground between synonymy and homonymy? The definitions of healthy in the different cases are clearly distinct (‘contributing to health [through nutrition]’; ‘preserving health [through excercise]’; ‘being in a state of proper functioning’), so they cannot be synonyms. But even though the definitions of healthy are distinct in the different cases, they are nevertheless related; they all make reference to health. Do they qualify as homonyms despite of this? In the first part of the thesis it is argued—based on textual support from the whole aristotelian corpus—that Aristotle holds that things having the same name and partly overlapping or related definitions also qualify as homonyms, and that there consequently is no third thing between homonymy and synonymy. A consequence of this is that all the so-called pollachos legomenon (‘things said in many ways’, some famous examples of which are: being, the good, nature, justice, friendship) are regarded as homonyms by Aristotle. After having outlined, in the first part of the thesis, the various types of homonymy recognised by Aristotle and having exhibited the way he uses his notion of homonymy both in destructive and constructive contexts (viz. both when critically confronting colleagues and busily building theory), the most philosophically interesting class of homonyms, the associated or related ones, are investigated more thoroughly. Due to the fact that the instances of this kind of homonymy all seem to associate around a core: the non-core instances inevitably making reference to a core instance, this kind of homonymy is labelled core-dependent homonymy. In this part of the thesis, Christopher Shields’ important and influential causal analysis of core-dependent homonymy is critically presented. In the third and last part of the thesis several challenges to Shields’ analysis are developed. Christopher Shields’ account has difficulties explaining the assumed assymetry in the dependence relation between core and non-core instances of homonyms, and when additionally some of Aristolte’s central examples of core-dependent homonymy seem to resist causal analysis, the tenability of Shields’ analysis is questioned. The last part of the thesis is concluded with a suggestion for an improvement of Shields’ causal analysis, as well as a brief sketch of an altogether different way of analysing core-dependent homonymy.