Situated atop a spur between the valleys of the River Monnow and its tributary, the Olchon brook, in the shadow of the Black Mountains, Longtown was known as Ewyas Lacy in the medieval period, after the Domesday holder of the estate, Roger de Lacy, whose main base was at Weobley; the first known use of 'Ewyas Lacy' (as a district) occurs in 1219. The region of Ewyas – an ancient name perhaps signifying a district of sheep-farming – was in Welsh hands in, and long before, 1066. Roger's father, Walter I de Lacy (d.1085) – given land at the Longtown site by Earl William Fitz-Osbern, and wider estates, held direct of the king, as a reward for his opposition to William's rebellious successor Roger de Breteuil – established there a Norman dominance that remained somewhat tenuous in his time and that of his immediate heirs. Although the nearby castle-town of Ewyas Harold – the castle perhaps marking the extent of Saxon penetration into the Ewyas region – was in other hands after the Conquest, the land surrounding it was held by the Lacy family, which is also associated with urbanization at Hereford and Ludlow. We hear in the Pipe Roll of 1186/87 of a "new castle" which is believed that erected at the site of Longtown, other Norman castles having been put up in the same area, including one at Walterstone, which could have been named for Walter I, and one less than half a mile from Ewyas Lacy at Pont Hendre, guarding a crossing of the Olchon; two castles in such proximity might point to conflict between two lords for control of the area, but any such struggle is dark to us, and it may be that one simply superseded the other, perhaps as late as the 1180s.

Roger de Lacy is generally credited with building either the Pont Hendre or the Longtown castle, though we cannot be certain; we can at least say the Lacy family emerged as possessors of both fortifications, with that at Longtown being the focus for upgrades and the Pont Hendre structure left to decay. Roger forfeiting his lands after rebelling against William Rufus, they were granted either to his brother Hugh, who had no sons, or to Gilbert, son of a sister of Roger and Hugh. Gilbert was succeeded by his sons Robert and then Hugh II de Lacy, who was lord of Ludlow in name at least, and Hugh by his son Walter II, who cemented an alliance with William III de Braose by marrying his daughter, in the period when William was on his way to becoming one of the leading magnates of the Marches.

The identity of the founder of a town at Longtown is, however, another uncertainty, the earliest documentary reflection of it being mention of a burgage in 1234. For one thing, Hugh II de Lacy had been murdered in 1185/86 and his son Walter II (born ca.1172) was, as a minor, unable to obtain control of Ewyas until 1189. During his minority, Ewyas was in royal hands, and we know that there was further expenditure on the castle; this direct royal administration of Ewyas was again the situation – further complicating the issue of identifying the town-founder – 1194-98 and 1210-13, the former occasion being when the king was concerned about Lacy ambitions in Ireland and the latter due to the Lacy connection with the Braose rebellion. If for no other reason, Walter would have been motivated to enhance his revenues, such as through instituting a town, because of the money he had to pay out to Richard I to recover his estates in England, Ireland, and Normandy. The castle had doubtless attracted settlement and was seen as a base for Walter to try to consolidate his hold in the region, against a constant Welsh threat. A town might have been founded once the castle had been built or, more likely, after Walter gained control of it in 1189. He could have taken that initiative following his restoration to favour ca. 1215 (again upon payment of a large fine), to assist the king against the alliance of rebellious barons with the Welsh ruler, he being appointed sheriff of Herefordshire in 1216. It seems less likely he would have founded the town at a later period within his lordship, as he was greatly occupied with the troubled state of Ireland in the 1220s, although he was continuing to build up sizable debts, and in his later years blindness and infirmity prevented him from being very active, except in efforts to fight off Jewish creditors.

Some small support for the interpretation that he founded the town at Ewyas Lacy in the earlier part of his adult life comes from the chapel serving the settlement (not a parish in its own right); although a thirteenth-century build, it incorporated a twelfth-century capital from an earlier structure, but we cannot be certain it was from the same site. This church was situated at Cloddock, a little south of where the borough would be established, and likely was an early focus for settlement. The borough's southern boundary adjoined that of Cloddock and was delineated by the Olchon and its continuation (from its junction with the Monnow), a stream called the Nant, while the eastern boundary was a ridge, up whose slope the borough extended, though probably only as fields or pasture. It has been suggested that the choice of the site for the borough was intended to minimize encroachment on established estates and thereby inflame hostility that could prove problematic to efforts to hold the frontier [Nina Wedell, "The origin and decline of the borough of Longtown", Ewyas Lacy Study Group, 2016, http://www.ewyaslacy.org.uk/-/Research-paper-The-origin-and-decline-of-the-Borough-of-Longtown/11th-to-19th-century/nw_lon_1036, last viewed 3 October 2017.].

No market charter is known for Longtown, but we need not expect one from a Marcher town of late twelfth century date. By contrast Walter II's son and heir Gilbert de Lacy acquired in 1220 licence for a market and fair at one of the most valuable of the family manors, Britford (Wilts.), while other branches of the family were rather more active in developing markets, albeit at a slightly later period and in other parts of the country. A nova villa within Ewyas Lacy is heard of in 1232 and reference to a burgage there about two years later; this was around the time of Gilbert's death, so that after Walter's death in 1241 the Lacy estates passed to Gilbert's two daughters, each of whom held a half-share in Ewyas Lacy. Over the previous decade Walter II de Lacy, whose stint as sheriff of Herefordshire lasted from1216 to 1223, was busy refortifying the castle at Ewyas with stone keep and bailey wall; this offers another context within which a borough might have been founded, although the absence of market licence would then be harder to explain. Longtown Castle remained into the early fourteenth century a valuable defence at a strategically important location – this strategic position, in terms of the Norman advance into Wales, was perhaps more important than the (limited) natural defences of the site.

By 1310 – a period of relative peace in the region – Longtown is recorded as having a hundred burgages. This is likely an approximation, but nonetheless suggests it was faring well enough as a small market town with a population of several hundred residents. An aerial survey has identified the outline of some thirty typical burgage plots, concentrated along the east side of the through-road just outside the defensive enclosure, while documentary surveys of early eighteenth century list just over two dozen properties owing burgage rents. However, Wedell [op.cit.] suggests that other properties held by burgage tenure may have been of atypical dimensions, including some that had just residences but little land to the rear, and others in which residence and open land were separated.

In 1271 the income from tolls was £20. The market is heard of again, along with a November fair, in 1369, in the inquisition post mortem on Sir Bartholomew de Burghersh (who must have come into possession of a share through his father's marriage to a Verdun heiress); by that date the market was a Thursday event. We have no further information on Longtown's economic history until the close of the Middle Ages, when it had ceased to be viable as a market centre. The lords who succeeded to the Lacy family estates – Genevills, Mortimers, Verduns, Burghershes, and Despensers – showed no more interest in developing Ewyas Lacy than they did Ludlow, apparently because their main interests lay in other regions of the realm, for the valuation of tolls in 1271 suggests there was sufficient local commerce to be warrant obtaining licence for an additional fair. That valuation may only have represented the Verdun share of profits, although this is not explicitly stated, as it is in a later inquisition post mortem (1316/17) when the Verdun share of tolls from market and fairs was assessed at a much lower £5, rents there also having fallen, although revenues from legal administration had risen by a third; we cannot place too much reliance on such evaluations, which tended to be estimations and affected by factors of which we are ignorant, along with those of which we are not (such as increasing toll exemptions). Another inquisition in 1331 valued the market toll at 60s. (without mentioning fairs), though the court profits continued to rise, while that of 1369 estimated a share of the fair (possibly the only one held there, as the plural was used to refer to a single annual event) at 40d. and of the market at 6s.8d. By this time the castle was being described as worthless (and in fact had been described as ruinous in 1328, not to the point of being unusable, but of costing more to maintain than it was worth), pointing to its neglect; yet the hundred court and hallmoot continued to do good business. These courts were essentially manorial institutions – a much-decayed court house (probably used by the lords of both moieties) still being visible in the sixteenth century, and court rolls still being compiled into the nineteenth – there being no evidence that the borough was ever granted any independent judicial jurisdiction that would warrant it having its own court. However, the burgesses seem to have had a measure of self-administration by the close of the sixteenth century, under a fee farm arrangement.

The almost rectangular castle enclosure is situated at the northern end of Longtown, with steep approaches to east and west, but level ground to its north. The castle layout is somewhat unusual because of the number of defensive enclosures – possibly originally outer baileys – surrounding it, which has led to speculation about re-use of an older fortification such as burh or Roman fort, though such archaeological investigation that has taken place has found no evidence to support this theory. Evidence from the defensive earthworks enclosing an area on the south side of the castle – generally interpreted as town defences rather than an outer bailey – suggest earliest urban settlement concentrated there. Fewer remains survive of something similar on the north side of the castle, and it has been debated whether that area was also enclosed, but it seems to have been on at least three sides. House platforms and walls have been observed in the northern section, some associated with twelfth and thirteenth century pottery finds. Most probably it was part of the borough, though whether of an original planted town or subsequent expansion is harder to say. The hundred burgages of 1310, assuming a standard plot size, could not have fitted within the town enclosure just south of the castle. Suburban expansion south of the original southern enclosure was itself given an enclosure at a date uncertain but probably around the heyday of the town ca. 1300; the regular layout of plot boundary earthworks here suggests this could have been a planned extension. Some population expansion may have been accommodated by settlement within what had once been an eastern bailey; but the east and west sides of the castle are slopes down to the watercourses and were used more for agriculture than human occupation.

This multi-unit layout stretching along the low ridge must have been what later gave rise to the name Longtown, first recorded in 1540. The north-south through-road (High Street) along the ridge passed through the centres of both the southern and northern sections of the settlement and along the western edge of the castle's eastern bailey; the masonry components of the refortified castle were all on the west side of this road. Not only the market area but the whole stretch of road passing through the town was much wider in the Middle Ages – the product of traffic meandering over an original rough track – than it is today; this was fairly common with major medieval urban streets that were used as marketplaces, and not for nothing are they often named Broad Street. The erection of the chapel across the eastern side of the road, at the castle end of the southern enclosure, made it easier for encroachments to take place further south along that side of the road. Since it sat in the road, the chapel is unlikely to have been surrounded by any churchyard.

The chapel stood on raised ground just south of the castle's older entrance (pre-dating Walter II's refortification), separated from the castle by an open area, and situated at the northern end of a widened, and roughly triangular, stretch of the through-road; this was clearly the marketplace, protected within the southern earthwork enclosure. The chapel's dedication to St. Peter could indicate an origin roughly contemporary with the marketplace. Much of this marketplace was later infilled, leaving the through-road on one side and a lane to the church on the other. Burgage plots were laid out: around the marketplace; perhaps on the east side of the road just south of the defensive enclosure; and along the east side of the axial street in the northern section of town. The west side of the northern street was occupied by tenements whose longer dimension ran parallel to the street, though these might represent post-foundation encroachments into the once-wider street. The cartographic record does not suggest the degree of consistency in plot size or shape that we typically associate with a carefully planned town, but the marketplace and earthworks certainly point to urban plantation. The northern section appears to have been depopulated, or even abandoned, after the thirteenth century.

During the Late Middle Ages Longtown was probably experiencing the same economic and population contraction as other Herefordshire towns. By the close of that period its market may have ceased to function, and an absence of post-medieval inns or lodging-houses, suggests a lack of visitation by outsiders, including those who would have frequented the market. In later centuries Longtown was considered simply a village. Restricted access to its geographical location may have made it an attractive site for a military fortification, but eventually told against its effort to develop into an economically stable market settlement.