In the 1870s, the Victorian antiquary G. T. Clark noted `some banks and ditches? in a field to the north-east of Coity castle. He interpreted these as possible siegeworks, but as depicted on the OS 25" 2nd edn map, they consist of a straight length of earthwork apparently a bank and ditch, with the beginnings of a rounded corner at the northern end, c. 100m north-north-east of the castle. The appearance of the earthwork, which had disappeared by the time the 3rd edn map was drawn up, is more consistent with part of the east side and the north-east angle of a Roman fort than with anything related to the castle. It lies on the line of the possible Roman road heading for the confluence of the Ogmore, Garw and Llynfi valleys a short distance to the north-west. The ridges between and to either side of these valleys are associated with a series of ancient ridgeways, and the castle, one of the major castles of Medieval Glamorgan, controlled these routes where they came down from the uplands into the coastal plain. A Roman fort at this site would have had the same strategic function. There have been no excavations either on or near this earthwork, nor in or around the castle.

E.M. Evans, GGAT, from the Roman Frontiers Publication, 2010.