One of the few castles built in England before the Conquest is believed to have been at Ewyas, atop a spur between two small valleys; along its east side ran the Dulas Brook. The castle is thought to have been built around 1049 by Osbern Pentecost, but dismantled after Earl Godwin was restored to power. Domesday Book states that William Fitz-Osbern, Earl of Hereford, rebuilt it, which must have happened soon after the Conquest, and put it, along with the surrounding arable, pasture, fishponds, tenant farms, and cottages into the custody of Alured of Marlborough. But by 1100 it was in the hands of Harold of Ewyas, son of the Confessor's nephew Ralph, a former Earl of Hereford (ca.1050s) with a Norman background. Around 1120 Harold's son, Robert of Ewyas, very active in the defence of the Marches, having enlarged the castle granted land south-east of it for relocation of a monastic priory from nearby Dulas into the better protection of the new outer bailey; it was in his interest to do so, quite apart from Christian motives, for the monks served the castle chapel and the parish church. However, the priory was beset by a variety of difficulties that brought it to poverty by the early fourteenth century and suppression in 1358; its presence would not seem to have contributed greatly to the prosperity of Ewyas. The castle passed, by marriage of an Ewyas heiress, to the Tregoz family in the early thirteenth century, and in the fourteenth into other hands, but no market or fair licence is ever recorded for it.

Domesday indicates that there were two messuages within the castle (bailey). Whether this was the beginnings of a town, as Beresford has speculated, or a pre-Conquest settlement known as Mulstonestone, for which there is slight evidence, remains uncertain; but the construction of the priory and later allocation to it of a close might have displaced older buildings beyond the barns that were said to have been taken down initially. Non-intrusive archaeological investigation of the castle site has identified numerous features within the outer bailey, which likely include seigneurial buildings, the priory, and village residences.

Later settlement, whether a relocation or not, was focused on the southern and eastern outskirts of the bailey, on the gentle slopes around the Dulas Brook. The context of transferring at least some of the residents of the outer bailey off the new priory site there would have been a convenient moment for introducing burgage tenure. Burgages are heard of at Ewyas from the late twelfth century, one of them being said to lie between the priory barn and the brook and apparently being the rear (curtilage) part of a burgage plot. In a detailed extent of the estate, made in 1300/01 following the death of John Tregoz without male heirs, Ewyas Harold is referred to by the ambiguous descriptor of villa, but there is an extensive listing of burgages, numbering about 70 (some sub-divided); their occupants – with both English and Welsh names – included a carpenter, turner, cooper, wheelwright, a saddler or packman, shoemaker, glover, three weavers (two female), a dyer, and two laundresses. At the same time we hear that two fairs were held annually. No market is ever referenced, but if the borough were an early foundation this would explain the absence of a licence under its early lords. Furthermore, the area was a poor one, agriculturally, and any market may not have produced sufficient profit to persuade later lords to purchase a licence.

A parish church was built, possibly in the late twelfth century (although rebuilt in the thirteenth), east of the bailey, on the far side of the brook. It stood, within an area of intensive settlement, not far from a junction of two roads, entering from north-east and south-east, with a street leading across the brook, near the site of a mill, into the castle. A wide area at this junction – with the brook on one side to serve as drainage – looks a suitable place for a market. The large number of burgages indicated ca.1300 would probably have ranged along either side of the through-road, near the church, although there is little trace of these in maps of the nineteenth century.

Above-ground traces of any medieval town at Ewyas have disappeared with the passage of time. The castle was ruinous by Leland's time, although parts were still standing. There is nothing, beyond the range of occupations evidenced ca. 1300, to indicate the borough was much of a commercial success, and it seems probable that the latter part of the fourteenth and the fifteenth century would have seen a process of decline. In the post-medieval period Ewyas was considered a village.