This article focuses on the properties of Old English clause structure in order to test Hinterhölzl’s (2014) hypothesis, namely that constituents at the left and at the right side of the VP surface pre- or postverbally according to the interaction between information structure and prosody: it is predicted that the relevant constituents may be spelled out either pre-verbally (movement) or post-verbally (in-situ), their mapping being driven by information structural and prosodic interface conditions. In this respect, I show that variation arises in Old English texts: Earlier stages of Old English show information-structural mapping of constituents, while prosody counted mostly for particularly complex constituents. By the 12th century, the prosodic option gains ground, favoring light constituents in pre-verbal position. The variation is compounded by an asymmetry between main and subordinate clauses, the former favoring a prosodic mapping of constituents. Key words: Old English, information structure, prosody.