This dissertation investigates the precise degree to which prosody and syntax are related. One possibility is that the syntax-prosody mapping is one-to-one (“isomorphic”) at an underlying level (Chomsky & Halle 1968, Selkirk 1996, 2011, Ito & Mester 2009). This predicts that prosodic units should preferably match up with syntactic units. It is also possible that the mapping between these systems is entirely non-isomorphic, with prosody being influenced by factors from language perception and production (Wheeldon & Lahiri 1997, Lahiri & Plank 2010). In this work, I argue that both perspectives are needed in order to address the full range of phonological phenomena that have been identified in English and related languages, including word-initial lenition/flapping, word-initial segment-deletion, and vowel reduction in function words, as well as patterns of pitch accent assignment, final-pronoun constructions, and the distribution of null complementizer allomorphs. In the process, I develop models for both isomorphic and non-isomorphic phrasing. The former is cast within a Minimalist syntactic framework of Merge/Label and Bare Phrase Structure (Chomsky 2013, 2015), while the latter is characterized by a stress-based algorithm for the formation of phonological domains, following Lahiri & Plank (2010).