The paper presents a set of data from Medieval Romance (MR), pointing out the relevance they have to the mapping of the left periphery of the clause (CP), building on the seminal work of Rizzi (1997). The phenomena presented here appear to support the proposal, first made in Benincà (2001) on the basis of modern Italian, that a lower section of the CP field only hosts moved elements, while higher sections host different kinds of base-generated topics; moreover, MR data can be exploited to shed light on other aspects of the left periphery and the constructions that are hosted there (such as questions and relative clauses). Data are taken from MR varieties, including medieval dialects of Italy, going from the 12th century to the early 14th century. The whole of MR languages are considered here a set of variants of an abstract Medieval Romance , characterised by important shared properties of the functional elements (V2 syntax and related phenomena). The primary focus will be the position of complement clitics in main and dependent clauses, where a set of generalisations will be established that can be stated for the whole MR. This crucially requires making reference not to surface positions, or to a course grained distinction between core and periphery of clause structure, but rather to specific functional positions in an articulated CP structure.