Extant accounts of scrambling in Ukrainian generally don’t extend past object- and other NP-related processes (Féry et al. 2007, Mykhaylyk 2010). Analysis of Slavic scrambling as XP movement (Corver 1992, Bošković 2005) runs into problems with split constituency, as does OT syntax (Gouskova 2001). Remnant movement (Sekerina 1997, Bašić 2004) runs afoul of Slavic data and theory too (Pereltsvaig 2008, Kariaeva 2009). Analyses that mix syntax with prosody (Antonyuk-Yudina & Mykhaylyk 2013; Mykhaylyk 2012) are more promising but also fail. Ukrainian scrambles only prosodic entities, ignores core principles of syntax, and respects core principles of phonology.The driving force behind the scrambling is not our focus here. It is generally assumed to be pragmatic in nature, based on things like topic, focus, and givenness (e.g., Féry et al. 2007). Fanselow & Lenertová have recently argued against this, however, and claim for most scrambling that ‘accentuation rather than informational status determines which categories can be fronted’ (2012:169); their findings support Chomsky’s (2008) view that information structure does not result in movement. We leave this to future research and focus here on which part of the grammar the movement takes place in. We propose that Ukrainian scrambling is phonological movement of exactly the sort found in Ancient Greek and Latin (Agbayani & Golston 2010, 2016), and similar to the more limited type found in Japanese (Agbayani, Golston & Ishii 2015) and Irish (Bennett, Elfner, & McCloskey 2016).