This article reports on the results of an experimental translation study conducted in 2008 in which 16 adult native speakers of the Croatian variant of Bosnian-Croatian-Serbian (BCS) were asked to translate nine texts from the Serbian BCS variant into their native Croatian variant in order to test the extent to which Croatian and Serbian do or do not employ distinct linguistic devices. The results show, on the basis of a statistical comparison of the purely grammatical building blocks in the original texts and their translations, that the Croatian and Serbian variants of BCS have essentially identical linguistic systems across all levels of language structure. In particular, we find that the phonological and syntactic systems are essentially identical and that over 98% of derivational and inflectional morphology tokens are identical. Lexically, the open classes show a difference of less than 10% of tokens, whereas the closed grammatical classes show identity in over 95% of cases.