Does linguistic rhythm matter to syntax, and if so, what kinds of syntactic decisions are susceptible to rhythm? By means of two recall-based sentence production experiments and two corpus studies-one on spoken and one on written language-we investigated whether linguistic rhythm affects the choice between introduced and un-introduced complement clauses in German. Apart from the presence or absence of the complementiser dass ('that'), these two sentence types differ with respect to the position of the tensed verb (verb-final/verb-second). Against our predictions, that were based on previously reported rhythmic effects on the use of the optional complementiser that in English, the experiments fail to obtain compelling evidence for rhythmic/prosodic influences on the structure of complement clauses in German. An overview of pertinent studies showing rhythmic influences on syntactic encoding suggests these effects to be generally restricted to syntactic domains smaller than a clause. We assume that, in the course of language production, initially, clause level syntactic projections are specified; their specification is in fact the prerequisite for phonological encoding to start. Consequently, prosodic effects may only touch upon the lower level categories that are to be integrated into the clausal projection, but not upon the syntactic makeup of the higher order projection itself.