Across a variety of languages, many words comprise more than one meaning unit, or morpheme. In the present chapter, reading studies employing readers’ eye movement registration are reviewed that examine how such polymorphemic words are identified in sentence context. The reviewed studies have examined how compound words, derived words, and inflected words are identified in sentence context. Studies are also reviewed that have investigated whether the meanings of polymorphemic words are constructed out of the meanings of their components. More generally, it is concluded that polymorphemic words are identified during reading using both whole-word representations available in the mental lexicon (the holistic route) as well as accessing the word identity via the component meanings (the decomposition route). Moreover, word length plays a significant role in modulating the relative dominance of the two access routes, with the decomposition route being more dominant for long polymorphemic words.