The present chapter describes word order variation in Romeyka based on the multilingual spoken language dataset of the Word Order in Western Asia (WOWA) corpus. Descending from VO ancestry, Romeyka shows under contact from Turkish increasingly head-final syntax. While cross-linguistically relevant factors such as semantic role, flagging and phonological weight do not offer much explanation for the flexibility between pre- and post-verbal constituents in Romeyka, information structure and phrase type do seem to be relevant. In addition, inter-speaker variation has been found significant to account for word order variation in Romeyka, suggesting that in a setting of language shift, individual forms of bilingualism affect word order.