In the following paper it will be argued that the phonological approach to clitic placement in Serbian/Croatian, as proposed in Zec & Inkelas (1990), not only fails to explain the observed phenomena, but also fails at the level of descriptive adequacy. Further arguments are presented against accounts which claim that clitic placement is syntactic and which utilize a post-syntactic operation of Prosodic Inversion (PI) in order to explain certain cases of apparent split of syntactic constituents (Halpern, 1992; Schtze, 1994). It will be argued that an alternative analysis which assumes syntactic clitic placement as proposed in Wilder & avar (1994) and avar & Wilder (1994) appears to be descriptively adequate. 1 Properties of Clitics in Croatian Table (1) gives a brief overview of diierent enclitic and proclitic categories in Croatian: (1) forms cliticization full reduced direction pronouns ACC fem: njj ju/njj msc: njega ga/nj auxiliaries pos jesam sam neg nisam prepositions nn nn ! The enclitic forms of the accusative pronoun for masculine and feminine diier depending on the syntactic context in which they appear. While the forms njj and nj are only licensed as complements of prepositions, i.e. only appear in prepositional phrases and only cliticize to a non-clitic (e.g. monosyllabic and bi-moraic) preposition, the other forms ju and ga may only appear elsewhere. 1 Another category that appears either as a full or as an enclitic form, is auxiliaries. Auxiliaries have two full forms, an aarmative and a negative form. The enclitic auxiliary appears in neutral contexts, i.e. neither emphatic nor negative contexts. Prepositions may be realized as independent words, if they are stressed (when they are bi-moraic, i.e. have a long vowel, for example in nn), or proclitic (when they are mono-moraic, i.e. have a short vowel, for example in nn). With the exception of prepositions and enclitic pronouns that only appear in PPs, all the other clitics, the reeexive pronouns and the question marker li form a morphological unit in which (apparently) the individual elements appear in xed positions: Earlier versions of this paper were presented at thè1996 Workshop on the Syntax, Morphology and Phonology of Clitics' at ZAS (Berlin), and at thè3rd International Summer School in Generative Linguistics in Olomouc'. I would like to thank all those who ooered their comments and criticisms, especially Maagorzata avar, Several informants consider the reduced enclitic pronoun in PPs archaic (e.g. Nedzad Leko (p.c.)), or even unacceptable (e.g. native …