During the late 1980s and early 1990s, the hypothesis of N-to-D movement played a prominent role in approaches to word order within the nominal domain. In particular, where N precedes all its arguments and modifiers, N-to-D movement was proposed to be the cause. Over time, descriptive and explanatory limitations of such accounts were discovered, and questions arose about the theoretical plausibility and desirability of head movement in general. Alternative approaches in terms of raising small NPs/XP were proposed to overcome these problems. In this chapter I trace the history of N-to-D movement: its empirical motivations, the bases for seeking phrasal movement alternatives, and flaws that have been pointed out in this latter kind of approach. I argue that XP movement competitors to head movement suffer from logical and explanatory problems themselves. I then make a case for N-to-D movement in the Shona language, where the word-order effects that it can capture correlate with inclusion of grammatical gender in all clause-level agreement. I argue that this is a result of adjoining N to D, and so is some loosening of the typical constraints on A-movement when N-to-D makes N's (uninterpretable) gender feature accessible to clause-level probes as an “activity” feature. An XP-movement approach does not connect these DP-internal and clause-level effects. I show also that a strict antisymmetric approach fails to capture simple regularities of word order in Shona DPs: DPs are noun-initial unless a demonstrative precedes N; otherwise, all modifiers are freely ordered postnominally. Symmetric base-generation options for modifiers is key to explaining this pattern, and N-to-D does the rest.