History of Wroxeter Roman City

Though a small village today, Wroxeter was once the fourth largest town in Roman Britain. It was almost the same size as Pompeii in Italy and its true size can best be judged by the scale of the surviving defences and the much-denuded ramparts. Founded in the mid-1st century AD as a legionary fortress, the town was established in the AD 90s, and inhabited until the 7th century. It is exceptionally well preserved – the site’s relatively remote location has meant that there has been little disturbance of the archaeological remains of the Roman town by later occupation. Archaeologists have been able to use cutting-edge technology here to reveal details of the city, which has led to a revolution in our understanding of this important site.

Aerial view of Wroxeter Roman City site
This aerial view of the excavated remains of Roman Wroxeter shows the basilica in the forground, the heated rooms and main bath suite to the left, the latrine area to the right, then the market hall, and the later western bath suite in the central courtyard

Pre-Roman Land Use

It is not possible at the moment to say when the Wroxeter area was first settled. Late Neolithic pottery and flintwork suggest people lived there or nearby about 5,500–5,000 years ago (3500–3000 BC) but we have no evidence for houses of this date here.

The geophysical surveys of the town have revealed a cluster of probable Bronze Age burial mounds (‘round barrows’) probably created around 1500 BC. A decorated fragment of an early Bronze Age ‘pygmy cup’ has been found in the town which may have come from one of these barrows and flintwork of this date has also been found.

It is possible that the barrow cemetery was established at a time when the banks of the River Severn were forming, as environmental archaeological evidence shows that the course of the River Severn was established at around this date when large amounts of silt were washed into the river and were rapidly deposited to form its banks. This silt probably entered the river system due to massive and sudden deforestation further upsteam.

If so it suggests that much woodland clearance of the lowland Severn plain was taking place at this time. From being a broad and shallow river, the Severn became a relatively narrow one that was difficult to ford.

The barrow cemetery may have been located in such a way as to create an obvious aiming point for the important ford at Wroxeter which today lies just by the church in the village. Despite this activity we have no evidence for people living at Wroxeter but clearly they were using the landscape here at that time.

The next major phase is the Iron Age, starting in perhaps 600 BC in this region. The evidence for this phase in Wroxeter comprises at least two ditched enclosures, one excavated underneath the legionary fortress established by the Roman Army in the late 50s AD and the other seen in geophysics.

Study of the hinterland around Wroxeter has indicated a broad pattern of single-ditched enclosures of this type in the landscape and, where excavated, these are Iron Age, Roman, or Iron Age and Roman in date. The widespread nature of these enclosures indicate a settled and farmed landscape in Wroxeter and its region with evidence for a mixed farming pattern of arable and pastoral agriculture set within a network of tracks and field boundaries. The Roman Army, when it arrived, thus moved into a settled and farmed landscape broadly similar to that which still exists today in the region.

One of the distinctive elements of this period is the relative lack of material culture found in these farms and settlements. The inhabitants do not seem to have used pottery in any large quantities, other than a coarse pottery containing salt imported from the brine springs operating in Nantwich, Middlewich and Northwich to the north of Wroxeter.

Metalwork is rare but often of high quality; an iron scabbard is known from an enclosure excavated beneath the modern Shrewsbury Town football club while elaborate bronze brooches dating from 300 BC to the 1st century AD are known from the region.

Coins do not seem to have been used in the Iron Age at all and it is unlikely that Roman pottery was being imported into the region, in contrast to the south and east of Britain where considerable quantities of Roman pottery were being imported and used before AD 43.

The inscribed tombstones of Roman soldiers buried at Wroxeter Roman City
The tombstones of Roman soldiers buried at Wroxeter; inscriptions record that soldiers serving at the Wroxeter fort came from France and Italy

Army Occupation (AD 47-90)

The Roman army first appeared in the Wroxeter region in about AD 47 when it advanced along the line of the modern A5, which largely follows the course of the Roman road, Watling Street. Another line of advance came from the fort at Greensforge in South Staffordshire so that both banks of the Severn were covered.

The legions appear to have stormed and set fire to the hillfort on the Wrekin which caused the tribal leaders to capitulate rapidly. The army then set about establishing a garrison in the region and continued its campaign to the west and north.

A small garrison of about 500 men was established in a fort 2.3 hectares (5.8 acres) in area on the east bank of the Severn 500 metres south of the village which was probably there to guard the ford. It may have held a detachment of cavalry since a tombstone of a member of a Thracian cohort of cavalry has been found.

Several other temporary forts or fortresses are known in the immediate area of Wroxeter suggesting it was a very active base but these were superseded in the late 50s AD by the establishment of the legionary fortress, the base of the 14th Legion, Legio XIIII Gemina and then the 20th Legion, Legio XX Valeria Victrix.

The fortress was located with its northern rampart on the highest point in Wroxeter on a gently sloping plateau between two streams running east–west that cut off approaches to the fortress from the north and south. To the west, the fortress was protected by the River Severn but to the east there was no natural defence.

The fortress was of conventional design with the main gate facing westwards, towards the enemy. We know little of the detailed arrangements inside the defences but parts of some barrack blocks have been excavated as well as some storehouses.

A granary is visible near a gate on the north side and we can guess the location of other buildings from the fact that Roman fortresses are often laid out in similar ways. Virtually all the buildings were of timber, as were the defences.

Phases of Occupation

Three phases of occupation have been suggested, more on the basis of what is known of Roman military activity in Britain at this period than from direct evidence in the fortress.

The initial phase of construction and use dates to about AD 57–66. The fortress was then refurbished for a second phase of use, AD 66–78, probably when the 20th Legion replaced the original garrison. Lastly, there is the final phase when the fortress seems to have been turned from an active base into a storage facility and reserve base.

It is possible that at this time the western defences were levelled and a further area between the fortress and the river was annexed to the fortress by cutting a double ditch from the north-west and (less certainly) the south-west corners of the fortress. This phase coincides with the campaigns of Agricola who took the 20th Legion campaigning deep into Scotland, only for them to be recalled in AD 86.

Initially the legion may have returned to Wroxeter where they perhaps started to build the bath house found unfinished in the annexe west of the fortress. Its construction was presumably abandoned by AD 90 because they had been transferred to Chester, levelling the fortress’s defences before they left.

Civilian Settlement

While the fortress was occupied a small civilian trading settlement seems to have developed close to the ford, south of the fortress. The finds – pottery from the Malverns and coins issued by the neighbouring tribe to the south, the Dobunni – suggests that the people who lived in this informal settlement may have moved to the region by travelling up the River Severn from the Worcestershire area.

Another, more formal settlement under the control of the army, may have been established north of the fortress, between the defenses and the lip of the Bell Brook valley.

The baths basilica, an enclosed exercise hall for the bathers in Wroxeter Roman City with a very high roof supported by columns
The baths basilica, an enclosed exercise hall for the bathers; the forum basilica, on the other side of the road, was equally grand © Historic England (illustration by Ivan Lapper)

The Roman Town

When the legions left in AD 90, the street grid and some of the buildings of the fortress were used to form the nucleus of the first town that replaced the fortress. This town was the seat of government for the tribal authority who now governed the region in the name of the Cornovii, the local tribe. Its town council will have included some veterans from the army but was probably mostly made up of local aristocrats.

They set out the boundaries of the town, taking in all of the fortress and its annexes, as well as both sides of the Bell Brook valley. A street grid of 48 town blocks was laid out, but not in an entirely regular fashion, as the local topography had to be respected. There were no defences initially and very few buildings of this phase have been recognised or excavated.

One building that probably does belong in this period lay south of the later bath house: this was the town’s ‘mansio’, a hostel where official travellers could stay and get a change of horses, a meal and a bath.

The newly created tribal capital was ambitious in size but in its early years appeared to have had no civic centre. That changed in the early decades of the 2nd century when a start was made on building the town forum and baths.

Civic Buildings

The founding of both baths and forum at the same time has suggested to some scholars that the Emperor Hadrian was personally responsible for ordering the construction of these buildings but there is no direct evidence to support such a claim, although it is a possibility.

The impressive inscription that had been located above the main entrance to the forum survives and gives us the date of its completion: AD 129–30, in the reign of Hadrian.

Archaeology tells us that the building had been begun about a decade earlier, and the foundations of the town baths had been laid at the same time. The bath complex, however, took much longer to complete, construction really only getting under way in the 140s with completion in around AD 150.

Both were laid out on either side of the main road through the town centre, which entered the town at its northeast corner and left at the southwest corner, where the ford was located. This main road was, in its turn, a new creation: the main road of the fortress had been located one city block further east. The new road line lay on the site of the western defences of the fortress.

One building missing from this complex of civic buildings was the town’s main temple. This was sometimes located within the forum, as is the case at Verulamium (St Albans), but no such arrangement existed at Wroxeter.

A colonnade of distinctive square-based columns found in the 1850s beneath the model farm complex north of the forum suggests that the town temple was located here. Certainly this would be a logical place to put such a building as civic ceremonies will have needed to progress from the temple into the forum, or vice versa.

Bath and Forum

The forum followed a standard plan for such buildings in this country: a range of shops along the street frontage, a courtyard behind with a colonnade around it on all four sides and, on the far side of the courtyard, a large basilican hall with a range of rooms behind it.

These rooms held the council chamber, the town’s main shrine, its record office and the offices of its two annually elected magistrates. The open courtyard held statues of civic dignitaries and, probably, of emperors, but was primarily used for a periodic market.

The baths followed a standard plan well known in the northern provinces of the Empire, where the climate demanded fewer areas open to the weather.

A long basilican hall, with its gable end facing onto the main street opposite the entrance to the forum, formed an exercise hall. At the east end of this, on the south side, a shorter arm of the building provided the main entrance (surviving today as ‘The Old Work’) and, beyond this, humid heated rooms like a Turkish Bath with symmetrical suites on either side providing a hot dry heat like a sauna.

The rest of the street frontage was supplied with two hot-food bars and a small square market hall (‘macellum’) where high quality food could be bought.

Housing, Defence and Industry

While very little of the town has been excavated, we can say quite a lot about how it was arranged and what kind of houses were built within it as the site is very receptive to various methods of remote sensing. The evidence from geophysical survey appears to show that the town was zoned into distinct areas of activity, as is found in other towns and cities of Roman date.

The stone houses of Wroxeter, for example, cluster in the centre, south and west of the town, where the prevailing wind would carry away the smoke generated by fires for cooking and heating and where the views to the west of the mid-Severn plain and Shropshire hills would have been uninterrupted.

In other words, the houses of the wealthy were located where one would expect them: in the most desirable locations. The limited number of houses that have been excavated confirm a mix of large town houses, some with evidence for mosaics and underfloor heating, and more modest dwellings including shops aligned onto the main street with accommodation behind.

At the highest point of the town, a large city block that is largely clear of buildings and is surrounded by a wall may have been the town’s livestock market. This interpretation is suggested by the fact that it is located in the highest and most exposed part of the town, where the prevailing wind would blow away the smell of the animals and where the town’s water supply, brought to this location by the V-shaped aqueduct, emptied into a huge cistern located just inside the town’s defences.

The defences themselves were built in the late 2nd century, in common with other towns in Britain, and comprised a turf and timber rampart, perhaps crowned by a wooden palisade.

This was refurbished in the 4th century, when the rampart was enlarged and the external double ditch was replaced by a single wider one.

The font in St Andrew's church, Wroxeter Roman City, is probably late Saxon or early Norman in date and it is made from the base of a Roman column
The font in St Andrew's church, Wroxeter, is probably late Saxon or early Norman in date and it is made from the base of a Roman column

The End of Roman Britain

Conventional Roman dating methods using pottery and coins are less reliable after the end of the 4th century as these items ceased to be imported or made in large quantities. The latest evidence of occupation lies directly under the ploughsoil and is thus not sealed by later activity, making it difficult to be confident of dating associations. For these reasons, great reliance is placed on scientific dates for the site in this phase, despite the imprecision that these dates provide.

Problems like these are not confined to Wroxeter alone: the end of Roman Britain and the beginning of the early medieval period is often characterised as a ‘Dark Age’ precisely because we can no longer be confident of what was happening.

Early excavators at Wroxeter conformed to a historical model which accepted that Roman rule ended early in the 5th century and that the end when it came was violent. The most famous example of this interpretation is the discovery by Thomas Wright of a skeleton with a coin hoard in the underfloor heating system of the main baths.

He assumed that this person had crawled into the baths to escape the sack of the town, evidence for which he saw in the black humic soil under which the ruins lay buried. We now know that the soil was black because it was rich in organic material rather than ashes and it is thought too that the body was put into the baths much later as a proper burial.

While scenarios like these may be fitting for towns on the eastern side of Britain, where Anglo-Saxon settlement was extensive and early, this is not the case in Shropshire where evidence for Anglo-Saxon settlement is rare before the 7th century.

More detailed and careful examination of the archaeology above the baths basilica has told a very different story in which the basilican hall and the baths were maintained probably deep into the 5th century and were then systematically dismantled.

Over the cleared ruins a number of building platforms, designed to support timber structures, were created, some quite substantial. These houses seem to have formed a nucleus of buildings clustered in the town centre: evidence for similar structures has been seen overlying the forum and above a town house immediately south of the baths.

While the structural evidence is reasonably clear, it is not clear how far buildings of this type and late date extended across the town, and whether such evidence represents the continuation of town life. Some aspects of this occupation suggests that town life was continuing, notably the evidence for markets and for the highly organised programmes of work that constructed these buildings in the first place, implying a hierarchical society.

If there were an aristocracy still in control, as seems likely, then their authority was presumably derived either from military or religious (probably Christian) authority.


Wroxeter’s abandonment as a settlement cannot be precisely dated but may have occurred in the mid sixth century, when a ‘Great Plague’ is known to have swept through Britain, or possibly in the seventh century, when the Anglo-Saxons took control of the region.

Although the town centre appears to have been quietly abandoned, people did not entirely leave the area. A small village grew up around the ford at the south end of the town, tucked in behind the Roman town defences and the natural stream that followed the same line.

The earliest unequivocal evidence of occupation is the Church of St Andrew whose north wall is clearly Anglo-Saxon in date, although it is built from reused Roman ashlar masonry. This building may date from the 9th to 11th century but built into the south side of the church is an earlier monument, an Anglo-Saxon carved cross of the time of King Offa (King of Mercia 757–96).

The existence of this monument demonstrates that Wroxeter was an important location at this time, and this is confirmed by the evidence that its parish uniquely straddles the River Severn, unlike all the other parishes along the Severn which belong on one bank to Hereford diocese or on the other to Lichfield.

Medieval and Modern Wroxeter

At the Domesday Survey (AD 1086), the church had four priests, again a sign that it was an important parish. It has been argued that this evidence suggests that there had been an existing British Christian community before the Anglo-Saxons built their church and that they perpetuated its importance.

The church was donated by Richard Fitz Alan to Haughmond Abbey in 1155 and, soon after, the church was enlarged, doubling its area. The church grew steadily through the Middle Ages, under the patronage of the Fitz Alans, The Earls of Arundel, who owned the manor of Wroxeter. The site of the manor house lay opposite the church where its extensive earthwork remains can be seen.

The church was much altered at the Reformation when the church tower was added, as was sculpture from Haughmond Abbey. The south wall of the church became dangerous in the mid 18th century and was rebuilt on a different alignment. Inside the church are a fine set of 16th and 17th century alabaster tombs commemorating the Newport family.

Of the medieval village we can say very little. We know that the focus of the village was opposite the church and a number of 16th and 17th century timber-framed houses survive. The surrounding pattern of open fields is recorded on an estate map drawn up by John Rocque in 1746. Shortly after this the land was enclosed, generating the field pattern visible today in the landscape.

The plan of the later medieval village is recorded on the tithe map of 1842 but the houses had largely disappeared by the time of the first Ordnance Survey map of 1880, its inhabitants no doubt moving away once the railways were built.

The only addition in this period was the model farm, built at the crossroads in the centre of the Roman town. Associated with this were a number of houses both next to the farm and in the village, and the main tenant’s house on the river cliff above the Severn.

About the Author

Dr Roger White FSA is Academic Director at the Ironbridge Institute and Senior Lecturer in Archaeology at the University of Birmingham. He worked as an excavator at Wroxeter from 1976 to 1990.

'step into englands story