Latest Project Publication in Science Advances

Our latest project publication in the journal Science Advances is now available open access.

Using cutting-edge isotopic science, applied to the archaeological site of Elverton Street, Westminster, it reveals this Tudor-era horse cemetery as the resting place of elite imported animals. Click this link to access it.

Read the University of Exeter’s press release here.

New Project Publication

We are delighted to announce our latest project publication, which you can read here.

This open access article in the journal Arms & Armour discusses the famous Warwick Shaffron, which is part of the Royal Armouries’ collection. Through a new programme of analysis, the article explores what this item and similar objects can reveal about the size of medieval horses.

End of Project Blog

This blog marks the conclusion of the formal funded phase of our AHRC Warhorse project, but our work of course continues as we develop publications and use the project’s achievements as a platform for future funding bids and other initiatives. So please do watch this blog for future updates.

The following account draws on the end of project statement submitted to our funders to first flag some of our key research achievements and, secondly, to reflect on how our research is having an impact on the wider world.

The headline is that our project succeeded in carrying out the first ever systematic and integrated study of the full range of archaeological evidence for warhorses and horses in medieval England — from scientific samples extracted from teeth through to analysis of bones, equestrian artefacts and armour, and the sites and landscapes associated with horse breeding and training.


Key Findings

The work has created new knowledge on three main fronts:

First, in the field of zooarchaeology the project applied cutting-edge GMM (Geometric Morphometric) and isotopic sampling methodologies to medieval horses for the first time. This has enabled the skeletal signatures of breeding and training regimes to be ascertained and for the biographies of individual horses to be revealed, and highlighted Europe-wide movements of breeding stock in a case study site in London. A headline finding about horse stature from a project database of 8500+ bones is that most medieval horses were below the size of modern ponies (1.48m to the withers, or shoulder), and detailed measurement of over 130 items of horse armour confirms the findings suggested from bone evidence. While popular perceptions of medieval warhorses picture large beasts, not a single bone from the entire project indicated a medieval horse that would have met the size criteria for a modern police horse (1.68m). Our zooarchaeology dataset is one of several important resources that we have now uploaded to the Archaeology Data Service website ( where they are freely available for researchers.

Image of the ADS homepage where datasets from our work can be accessed and downloaded.


Second, in the field of landscape history, the work has generated entirely new information about an aspect of the medieval countryside (the horse stud) that had been neglected. Over 800 medieval documents have been examined to provide information on the nature and organisation of studs, principally during the High Middle Ages. The project produced the first ever mapping of the English royal stud network, clarified the number and location of stud sites (within parks), characterised their landscape contexts (frequently in uplands or wetland zones), and reconstructed their hinterlands. The work highlighted that while studs can be located to places, physical traces of their buildings and infrastructure are exceptionally ephemeral.  It has also drawn attention to the tendency to misidentify equine structures, especially stables, as being for other livestock and purposes.

Third, in terms of debates within medieval archaeology and history the project has provided new evidence for continuity and/or change across the Saxo-Norman divide by highlighting that in terms of equestrian material culture the Norman Conquest had a minimal signature, with the introduction of the curb bit the principal indicator of change. Mapping of equestrian material culture from a project database of 14,000+ separate artefacts shows great potential, previously unrecognised, to map the footprints of medieval tournament sites and (through harness pendants decorated with arms and symbols) noble families. Follow the ADS link above to access our online dataset.

As an overall methodological achievement, the project’s integrated approach has made a powerful statement about how barriers between traditionally discrete research fields can be broken down to produce new understandings. In answer to the key questions of (1) whether we can identify warhorses in the archaeological record; and (2) whether the warhorse was a breed, a selection or some sort of cultural product, we can reach some preliminary conclusions by combining all our lines of evidence. First, it is rare for warhorses to be deposited in archaeological contexts which clearly identify the function and role of the animals. It is, however, clear that exceptional care was taken to breed and train appropriate horses, including importing or being gifted suitable animals from overseas. Whilst selective breeding in a stud network formed a key part of warhorse management, the destrier cannot be considered to be a ‘breed’. It more likely represents a selection for appropriate temperament and body conformation, accompanied by intensive training that resulted in highly valuable steeds. Warhorses might additionally be marked out to some degree by activity-related bone remodelling and pathologies. While the ‘great horse’ was likely of a larger stature than the average medieval equid, the metrical evidence from animal bones, horse armour and other material culture, suggests that raw size was considered less of a key trait than the physical and mental characteristics highlighted above. 


The impact of the project on wider non-academic audiences and stakeholders was of central importance to our work. The key areas of impact, recorded through the period of the funded project (Feb 2019-March 2023), fall into four areas.

First, the project generated benefits for our partner organisations. The Portable Antiquities Scheme, run through the British Museum, benefitted from the delivery of three intensive training workshops on recording equestrian artefacts to over 100 staff and volunteers and from the delivery of 10 freely available finds guides to help metal detectorists record items such as harness pendants and horseshoes (see According to the Head of the Portable Antiquities Scheme: “What is most important is that the work of the results of the project’s partnership working – the training workshops and freely-accessible finds guides – have a legacy, transforming the way that staff and public volunteers record metal-detected equestrian objects” (Professor Michael Lewis, 2023). The Royal Armouries benefited from the experience of working with an academic partner, including enhancement to its catalogue and records, opportunities for the co-publication of results, and enrichment of outreach activities (the project had a presence at two public events run by the Armouries attended by over 17,000 members of the public).

Second, the project is playing a key role in changing public perceptions about the size and appearance of medieval horses, which are often based on experiences of film and TV.  A press release which accompanied a project publication in The International Journal of Osteoarchaeology generated 197 press articles in 34 countries, while project staff have contributed through the podcasts and programmes including BBC Radio 4 ‘In Our Time’ to reach mass public audiences. The project had a presence at five large public events attracting 41,400 members of the public in total and generating reactions from participants such as “I had no idea a medieval warhorse could be so small; it’s fascinating to hear that historical films can get it so badly wrong!”

Third, the project’s findings have been actively promoted and disseminated on the national and international conference circuit, with staff speaking at 18 separate events in the UK to reach in excess of 750 participants and 15 events in a further 10 countries reaching a minimum of 550 participants.

Compilation image of outreach events and lectures involving members of the Warhorse team


Fourth, the academic impact of the work is having knock-on effects for the way that equestrian heritage is presented in museums and heritage sites. For example, Historic Royal Palaces reported in January 2023 that their website for Hampton Court has been amended to reflect our project’s findings. Another example is the way that English Heritage’s newly installed signage and the new (2022) official guidebook for Launceston Castle, Cornwall, now recognise a building on the site, which was previously misidentified, as a medieval stable, on the basis of the project’s findings.

Beyond these impacts, the high level of public interest in all manner of equine pursuits, such as show jumping and horse racing, is likely to open up opportunities for further aspects of public engagement, which the project will exploit, including through a Follow-on-Funding Bid.

With so much to shout about, and such exciting possibilities for the future I can only end this blog with the most enormous thank you to the incredible, energetic team of talented researchers that have driven the work forward, as well as to our project partners and everybody who has shown an interest. But the work continues…  

A good horse has no colour … or does it?

In this blog post, Dr Gary Baker shows how the documents for England’s royal warhorses can be used to show us what they looked like.

One of the most interesting facets of English medieval documents concerning royal horses is that many of them provide details of what the animals looked like in terms of their colouration and distinguishing marks. This is information that is often sorely lacking for medieval people and shows just how important horses were in the medieval world.

The colouration of horses is revealed by their names. In English medieval documents concerning royal warhorses (all of which were male), the naming practice is always the same. A horse’s “forename” was its coat colour expressed as a Latin or French noun. There were around ten to fifteen types, of which bausan (white spots on a coloured ground), bay/badius/bayard (bay), ferrandus (iron-grey), grisel (grey), morel (shiny black), lyard (very small spots of white on grey, or vice-versa, giving a silver-grey appearance), and sorel (chestnut) were the most common. The horse’s “surname” is either that of the former owner (surname or title) or the place from whence the horse came either by birth or purchase. Occasionally a further description is appended, describing some other physical characteristic. This is most often related to the coloration of the feet or head if it differed from the rest of the animal (like a white nose), but it can also refer to some other corporeal trait like a docked tail.

excerpt of a medieval manuscript

A typical list of the names of the royal horses

To provide an example of how historians can use this information, let us take a small sample size from the accounts of John Waterton, Master of the Horse (the chief officer in charge of equine matters) under Henry V. The most prevalent coat colour for the royal horses in Waterton’s account for 1414–16 was “lyard”, or silvery-grey horses. Of the 233 horses in the Master’s care over these two years, ninety-eight were silver-greys with the second most common being bays at fifty-five. If we include the silver-greys with horses described simply as “greys” (grissel) this number rises to 117, or half of all the king’s horses in the account.

pie chart of horse colours mentioned in the English royal stock accounts

Horse coat colours in the English royal stock 1414–16

Interestingly, a good proportion of equine literature produced in Western Europe during the Middle Ages had a good deal to say about the conformation and personality desired in the “ideal horse,” including warhorses, and many also remarked on desirable coat colours. One of the most influential medieval works was De medicina equorum (c.1250–56) by the Italian knight Jordanus Rufus which survives in numerous contemporary copies in several languages. In one of these French, or rather Occitan copies, Lo libre de la marescalcia des cavals, dating from the late fifteenth century, now in the Bibliothèque Municipale in Fréjus, France, “Rufus” states that there are so many different opinions about the optimal coat colours that it was not worth discussing them all. This is illustrated by divergences at this point of the text in several of the French versions of Rufus’s work, where the copyists express their own opinion as to the optimal colour for a “good” horse. For example, the Fréjus MS says that bay is the preferred colour over all others while a copy in Reims by François du Tronchoy, Livre de la cure et garde des chevaux (1390) states that bay and grey are the most advantageous. The evidence from Waterton’s account of the predominance of gifts of bays and especially greys made to, and by Henry V, constituting 109 out the 135 received by the king (80 per cent) and thirty-five of the fifty-two given away (67 per cent), suggests that ideas of equine colour preferences, or at least the prestige of animals of these colours, had taken root by the time of Agincourt. It is telling, for example, that the famous image of Sir Geoffrey Luttrell mounted on horseback in the Luttrell Psalter (c.1325–40) shows Geoffrey mounted on a dappled grey warhorse.

Sir Geoffrey Lutterell assisted by his with and daughter-in-law (British Library Add. MS 42130 f. 202v.)

Dark Horse – The Warhorse Project at the EAAs in Budapest

Our team has been busy investigating medieval (war)horses since 2019 and it was time to share our work with the research community on the international stage. So, we headed to Budapest (in person and digitally) to host the session ‘Dark Horse – Archaeologies of Horses and Horse Culture in the Middle Ages’ at this year’s European Association of Archaeologists (EAA) Annual Conference.

Oliver Creighton introducing the Warhorse Project

In a lecture room on the campus of Budapest’s Eötvös Loránd University, we attracted a sizeable crowd very interested in all the horse-y research presented throughout the day. Colleagues from around the world showcased their explorations into medieval equine archaeology – we heard about various aspects of equestrian material culture, such as horse armour, bits and harness pendants; how horses were treated in life and death; and several papers asked what a medieval horse (ought to have) looked like. Interdisciplinarity was clearly evident in all of the presentations; from history and traditional zooarchaeology to genetics and isotopic analysis, everyone made valuable contributions to the bigger picture (see here for the Session line-up).

We started off with an introduction to the session and the Warhorse project by our PI Oliver Creighton who highlighted how we are interrogating multiple lines of evidence – from horse remains through to artefacts, landscapes, documents and buildings – to create new understandings of the equine world in medieval England. ‘Zooming’ in from his professional-looking recording studio at the University of East Anglia, Robert Liddiard explored the historical context for horse breeding in the first half of the 14th century and revealed how place names can help identify key places within the royal stud network. Camille Vo Van Qui, staying on the historical side of things, gave an insight into the contemporary perception of the ideal (war)horse through Jordanus Rufus’ well-known manuscript De medicina equorum. Many participants reflected on how his descriptions of horses shine a different sort of light on the zooarchaeological evidence.

Kate Kanne presenting her work on (war)horse pathologies

The practicalities of certain conformations and their ability to perform when ridden were at the centre of Markku Niskanen’s talk. His findings aligned intriguingly with our own results based on the biometrical analysis of medieval horse remains, which were presented by Helene Benkert. Her PhD examines changes and developments in horse size and stature across medieval Europe and the preliminary results confirm the predominance of pony-sized horses we proposed for England. As per their occupation, medieval warhorses were prone to injuries and other pathologies. Using some stunning imagery, Kate Kanne highlighted how this line of evidence may help in identifying warhorses but also other types of horse usage in the zooarchaeological record.

Using the case of medieval Brno (CZ), Vera Klontza-Jaklova presented her full-suite analysis of the horse skeletons which had been unusually disposed of in shallow pits in backyards. Sporting a variety of shapes and similar average size, these horses fall well in line with trends observed in the rest of Europe. In Poland, Daniel Makowiecki informed us that the historical information on horses is equally scarce as it is in England, despite the animal’s great importance. Interestingly, evidence for hippophagy suggests that the Slavic population consumed horse meat on a regular basis, a clear taboo in much of Western Europe!

Roman Shiroukhov took us to the Sambian Peninsula in Kaliningrad Oblast, Russia, to explore the distinctive horse burial rite in early medieval Prussia. He supported his investigations with stable isotope analysis and C14 dating to further our understanding of those horse burials or sacrifices. Returning to medieval Poland, Danijela Popovic presented the genetic aspect of the interdisciplinary project ‘Horses in Poland in the Early Piasts and Internal Fragmentation’. In this context, medieval equestrians seem to have had a fable for chestnut horses and, as in much of Europe at the time, stallions (or geldings).

Carly Ameen remotely presenting her research on GMM

Continuing our exploration of horse conformation, Carly Ameen introduced us to her ground-breaking work on geometric morphometric analysis within the context of the Warhorse Project. She hopes to use GMM to detect in-life activity-related remodelling and usage-related conformation in metapodial and tarsal bones in horses. Alan Outram informed us how horse armour and horseshoes can contribute to the conformation debate, while discussing the problems posed by object design. Measurements taken from late medieval and early post-medieval shaffrons (head armour) as well as horseshoes largely confirm the average size estimates for the faunal remains. Our final speaker, Robert Webley, presented his work on equestrian material culture, specifically stirrups and bridle bits, highlighting its potential to connect iconographic, historical and archaeological evidence to better understand medieval equestrian culture.

Since Ludovic Orlando was, unfortunately, unable to join us and present his large set of genetic data we had some time to hear about Anna Szencsenyi-Nagy’s exciting new project for which she had prepared a poster. The project aims to explore how horses and humans interacted in Avar-period Hungary, using a variety of methods from traditional zooarchaeology to genetic analysis. We finished off this excellent session with a final round of discussions before relocating to the campus bar for some well-earned refreshments and more, friendly, debating.

It was a pleasure to hear about all this new and exciting research and to see the puzzle pieces fall into place with each new presentation. We would like to extend a massive thank you to all our presenters and the attentive audience for their interest, engagement and fantastic contributions. We greatly enjoyed the session and are excited to hear more about all the excellent work that was presented!

Part of the team and a couple of Exeter PhD students with their well-earned post-conference drinks. Cheers!

PODCAST: Future Imperfect – How big were mediaeval warhorses?

In this podcast, Oliver Creighton and Alan Outram discuss the nature and characteristics of the medieval warhorse at length with Jason Kingsley. Jason has a wealth of experience in riding horses fully armoured and training them to do so. It makes for a fascinating three-way discussion, weaving together their different areas of specialist knowledge.

To listen to the podcast click here.


The Warhorse Project at Tewkesbury

After three years away, the Tewkesbury Medieval Festival was finally back, running across the weekend of the 9th-10th July. We were back too! With a boot full of (horse) skulls, posters and horsey bits and bobs we headed up to Tewkesbury on the Saturday morning to present our research to the re-enactors and visitors of Europe’s largest medieval re-enactment event.

Sarcophagus of Sir Guy Brian

Tewkesbury is a beautiful place, and well worth a visit, especially the abbey which is a stone’s throw away from the field where the Battle of Tewkesbury took place. The space is serene now, but was not so in the aftermath of the battle in 1471 in which so much blood was spilt within the abbey’s walls that it required reconsecrating. Today, it contains the tombs of many notable medieval people including George, Duke of Clarence (brother of Edward IV), Eleanor de Clare, Hugh Despenser the Younger, and many other earls and knights (though sadly with no images of their warhorses that we could find)!

As usual, it was sweltering the entire weekend, but that did not seem to stop anyone from attending the festivities and market. Luckily, we had a stall in one of the marquees and could escape the sun, if not the heat. Many people stopped by to learn about medieval horses and their use in warfare, and some even remembered us from our very first and much smaller appearance pre-pandemic! 

Our two mares – one in a fitting shaffron, the other in a slightly too large bit, demonstrating where in a horse’s mouth it sits and how it works.

Outside the marquee we unveiled our new poster which invited people to guess the height of a medieval warhorse. The craft corner was a favourite with the younger ones, who replicated harness pendants with great enthusiasm, and many a visitor was fascinated to hear about the research our project has undertaken, from both a material and historical perspective. Our two lovely mares turned many heads and patiently helped visitors discover their species’ secrets. Seeing their live counterparts fully caparisoned during the re-enactment of the battle was a wondrous sight to behold. 

If you are curious to find out what sort of things were fed to warhorses in the Middle Ages, have a look at our flyer from the event and see how many you can get right (the answers are at the bottom – we won’t tell anyone if you sneak a glance)!

We chatted with many lovely people over the weekend, from interested members of the public to keen equestrians and (former) professional jousters. We also had the chance to meet Zac Evans, modern jouster and YouTuber, and filmed a short reel with him discussing our project and popular misconceptions about medieval warhorses. You can see the video here.

All in all, it was a fantastic weekend outing and a delight to see so much interest in our work. We plan on a few more of these events in the future so watch out for news on our Twitter feed and website.


Some impressions of the weekend:

The Size of a (War)Horse

Did you know that medieval horses, on average, were no larger than modern ponies? 

When thinking about ponies many people seem to imagine the smallest of their kind, something akin to a Shetland pony. But ponies are much more varied than that. The term ‘pony’ is an early modern invention and first appears in the mid 17th century, probably derived from the French poulenet meaning ‘small foal’. Nowadays, the Fédération Equestre Internationale (FEI) defines a pony simply as a horse that does not reach more than 1.48m (c. 14.2hh) at the withers.

Outline of a horse showing how the withers heights are measured at the top of the shoulder.

Withers heights are measured from the ground to the withers, the highest point at the shoulder. Technically, we are not measuring shoulder height but rather the top of the processes of the spine between the shoulder blades. HH stands for ‘hands high’ and is the common measurement for horse height in the UK.

That still sounds pretty small? Well, we recommend getting up close with a pony and see for yourself: they can be big, powerful animals! A modern Riding Pony for example, reaching 1.44 m (c. 14.1 hh) on average, weighs somewhere in the region of 390 kg (c. 860 lbs). Now imagine even just a few of them in full armour and with equally armoured and armed riders on their backs, charging at you. That must have been quite a sight.

silhouettes of a human, a horse and a pony next to a scale to show relative sizes.

The sizes of an average modern sports horse and a pony compared to an average modern man.

Although medieval documents are notably silent on the matter of size, we know from the zooarchaeological record that horses in the Middle Ages mostly stood at 13 – 14 hh at the shoulder (see here for our study). Only very few animals reached withers heights beyond the pony threshold. It is only in the post-medieval times that larger horses appear more frequently and much later still when we first see horses of the size of modern sports horses and heavy drafts. 

a box and jitter plot showing horse size variation and change from the Roman to post-medieval times and in comparison to modern specimens of different breeds plus donkeys and mules.

The change of withers heights in horses through time compared to modern animals (in grey). Each dot represents an individual within its population or time period, the thicker black line in the box shows their mean. Even the post-medieval mean is still within pony range but the overall range of the horses has increased and we see, for the first time, horses being regularly taller than 14.2 hh.

An extract from the Bayeux tapestry showing several riders with their legs hanging low, indicating smaller horses.

The famous Bayeux tapestry shows, like other medieval depictions, riders with their legs hanging below their mounts’ bellies. It confirms our archaeological findings that medieval warhorses were not the huge heavy beasts that people tend to imagine them as.

There are advantages to a smaller mount, especially in a warfare scenario. One of the more obvious ones is “mountability”. If you are unseated on the battlefield a smaller horse will make remounting, in armour, much easier. Big horses can often be a bit uncoordinated with their long limbs and may be more difficult to ride in good posture. 

As with many species, it is the small ones that can be the most ferocious. Perhaps this made the medieval ponies especially suited to warfare. Historic documents instruct the medieval horse trainer to encourage aggressiveness in warhorses and there are even illustrated manuals that detail kicking and biting manoeuvres.

In the end, the size of a warhorse did not really matter all that much in the Middle Ages. Medieval writers didn’t even find it noteworthy, preferring instead to detail financial values, age, colour, origin – those were of more importance to the people dealing with horses. And their abilities, of course! Those that went on to become warhorses were carefully chosen, not based on size but their conformation and personality (check out our work on the royal stud at Odiham here), the life of people and success in battle depended on their skill and suitability. 

An Archaeological Black Hole? Medieval Stables and Tattershall Castle

For all the importance of horses during the Middle Ages, their dedicated buildings are often elusive. Nothing illustrates this more clearly than the subject of stables. Even at major medieval residences it often requires considerable effort simply to locate where these buildings were on the ground. Such is the absence of physical evidence that the medieval stable has, not unreasonably, been described by Giles Worsley as ‘an archaeological black hole’.

Tatterhsall Castle, Lincolnshire. Ralph Lord Cromwell’s Great Tower. The potential stable can be seen in the background.

The lack of evidence is a result of several factors. Historically stables tended to be constructed in the outer baileys and enclosures of castles and palaces, areas that have not tended to form the subject of historical enquiry or prioritised for direct archaeological intervention. At the same time, many stables would have been timber-framed buildings and, like so many others, failed to come down to us as a result of fire, dilapidation, or demolition. The small number of physical examples that do exist survive in a heavily altered state and are frequently labelled simply as ‘ancillary buildings’ that have rarely been the focus of close study.

The castle’s ‘Great Stable’. Or is it?

But as the research from ‘Warhorse’ is showing, an additional reason why the medieval stable is somewhat shadowy is because of misidentification. Stables are sometimes hiding in plain sight, erroneously labelled as halls or barns, while repurposing in the post-medieval period has meant that others have in fact survived but with their original function unnoticed. As buildings, medieval stables often have more complex histories than might otherwise be thought and unravelling these complexities in specific cases is key part of our research.

With this issue in mind, ‘Warhorse’ commissioned Dr James Wright from Triskele Heritage to undertake a building survey at Tattershall castle in Lincolnshire. The castle is well-known as the home of Ralph 3rd Lord Cromwell and Lord Treasurer of England who transformed the site from the 1430s via an enormous programme of building works. Even allowing for early twentieth-century restoration, how the castle’s Great Tower would have dominated the skyline of this part of medieval Lincolnshire can still be gauged today.

James at work.

The reason for the project’s specific interest in Tattershall is that the castle’s partially surviving building accounts refer to the construction of a ‘great stable’. This building evidently attracted considerable investment as the accounts describe foundations, the use of several hundred thousand bricks, the existence of a paved floor and possible reference to some kind of upper storey. Previous scholars have identified this documented building with a large, ruined structure in the castle’s outer ward. Here archaeological excavation has found evidence for an interior drain running the length of the building and tethering rings still survive on one of the exterior walls. The existence of the footings of a smaller building to the northeast also seem to tie in with a detail given in the historical accounts that the great stable lay to the west of a woolhouse. Even though it exists in a fragmentary state, any opportunity to learn more about the physical nature of a medieval stable, especially one at a castle and in turn one that is documented is welcome. Cue James, some photogrammetry, and a long hard look at the surviving remains.

Without anticipating the results of the survey and as we rather suspected, the Tattershall ‘stable’ may not be all it seems. There are certainly some oddities that require explanation, not least that the spinal drain runs directly into the base of a fireplace in the end wall. There are also the remains of a wall that formed an internal partition, but this is of stone construction and as the rest of the building was built in brick it is suggestive of a later alteration. The tethering rings were clearly inserted into the main walls and so not part of the original build. Finally, at only 1.44 metres wide, the original entrance to the building seems rather too narrow for use by horses. Certainly we are looking at a building that was used as a stable for part of its occupational history, but it may not actually be the ‘great stable’ mentioned in the castle’s building accounts. Did the archaeological black hole just get a little darker?

Tethering ring on exterior wall. And don’t climb on the walls remember.

Well, if this building turns out not to be what scholars have hitherto believed it to be, then we can on the one hand perhaps forgive ourselves a little disappointment. But on the other, it would actually be rather illuminating. Later in the fifteenth century and into the sixteenth, the history of the castle is one of reduced occupancy following the death of Lord Ralph. The need for stabling would have consequently diminished and so we should be alive to the possibility that we might be looking at a building that was converted for use as a stable, perhaps from a range of lodgings. Adaptation from lodgings to stables (and indeed visa-versa) is something that we suspect happened at other sites and so Tattershall might end up standing as a well-documented example of a phenomenon that finds analogy elsewhere. And, after all, we might still have a medieval stable, albeit one of slightly different date than previously thought.

If, in the final analysis, our building turns out not to be the ‘great stable’ of Cromwell’s accounts, then where was it exactly? Somewhere in the inner or middle wards of the castle would seem most likely (it was described as ‘within the castle’ in the 1430s), although the precise location remains unknown. For now at least, the black hole will have to retain some of its mystery.

Huge thanks to The National Trust staff at Tattershall and Rosalind Buck for granting permission for the work to take place and to James not only for all his efforts but sharing his encyclopaedic knowledge of the castle with the project team. All information correct at the time of blogging!


Post by R. Liddiard

What were medieval stables like?

In this post, Dr Kate Kanne details the Warhorse team trip to medieval stables.

In an effort to further understand how medieval and Tudor horses were husbanded and utilised, and as a good excuse to enjoy the glorious spring weather, the Warhorse team recently visited some of the earliest stables built in England that are still standing. Though largely post-medieval in date, the stables at Kenilworth Castle, Warwickshire, and Dunster Castle, Somerset, can provide details to flesh out what the archaeological and historical records cannot. 

Prof Alan Outram and Prof Oliver (Ollie) Creighton were accompanied by myself, Helene Benkert (PhD student), and Tess Townend (MSc student). Alan and Ollie’s deep knowledge of medieval castles, horses, and their archaeology was augmented by our practical experience working with horses in a variety of equestrian disciplines in many kinds of stables. My professional equestrian background is primarily in polo, where large numbers of horses have to be fed, tacked, exercised and competed in sensible ways from a labour standpoint that respects the needs of equine athletes. I have co-authored educational materials for new horse owners on how to house and manage horses (Brady, Kanne, and Russell 2002a, 2002b). The organisation of medieval stables is familiar to modern equestrians that deal with how to best manage working horses used on a daily basis.

Kenilworth Castle was built as a Norman stronghold and modified over many centuries to form a spectacular residence (Fig. 1). The stables now standing (Fig. 2) were originally built by John Dudley, Duke of Northumberland and Master of the Horse of Edward VI, in 1553, at a cost of £4110, likely on the site of earlier stables. After John Dudley’s execution, his son Robert Dudley, Earl of Leicester, made additional improvements to the stables, also called Lord Leicester’s Stables. Like his father before him, Robert was the Master of Horse for the monarch, in this case Elizabeth I, with whom he had a special relationship. 

Kenilworth Castle and Tiltyard for jousting.

Fig. 1. Kenilworth Castle with the tiltyard for jousting in the foreground. Photo by Helene Benkert.

Kenilworth Stables

Fig. 2. Kenilworth Stables. Photo by Oliver Creighton.

Dunster Castle was constructed shortly after the Norman conquest by William de Mohun, initially as a fortress of earth and timber. The Luttrell family purchased the castle in 1376 and created the standing mansion in the early 17th century (Fig. 3). Stables are known from the 16th century, although the impressive example that can be seen at Dunster today was built in the 1660s, as confirmed by dendrochronological (tree-ring) analysis of oak beams that replaced those destroyed in the Civil War (Howard et al. 2003). Situated within a two-storey barn, the stalls date from the late 18th and early 19th century. In 1925, the Maharaja of Jodhpur visited Dunster Castle to play polo (Fig. 4), with a string of 62 ponies that were stabled there, with a groom for each pony. A video was made of the event (link here).

Dunster Castle

Fig. 3. Dunster Castle. Photo by Oliver Creighton.

Polo at Dunster Castle

Fig 4. Polo at Dunster Castle, Rao Rajah Manut Singh centre.

Tie, or standing, stalls were the norm until very recently for working horses around the world, including royal stables such as the Royal Mews at Buckingham Palace (Figs. 5 and 6). They require less space than box stalls, with the typical rule-of-thumb being 4-5 ft wide x 9-10 ft long for average sized modern horses (15-15.2 hh) (Fehr et al. 1993), and require less labour to muck out each day. Bedding would have likely been straw, like at the 1920s Royal Mews below (Fig. 5). Tie stalls make much sense to house the riding horses on the castle grounds, both for residents and visitors. They still are used in polo and military stables as they are useful to ensure each horse receives the correct ration of feed for its workload and body condition, and are handy for horses that may be used every day, repeatedly throughout the day, or at a moment’s notice, quickly available for grooming and tacking up. 

Tie stalls at the Royal Mews, c. 1920s

Fig. 5. Tie stalls at the Royal Mews, c. 1920. Photo by The Royal Collections Trust.

Tie stalls at the Royal Mews today

Fig. 6. Tie stalls at Royal Mews Today. Photo by The Royal Collections Trust.

Horses generally can lie down in them despite that they are often termed ‘standing’ stalls. They can either stand facing the wall with a hay rack or manger along the wall, like at Dunster (Fig. 6), tied to the wall usually with a rope and weight pulley system, and/or be cross-tied after feeding at the front of the stall facing the aisle as shown in Figs. 5 and 6 above. Though horses can fight when stabled next to one and other, especially stallions, solid partitions, or half-solid/half-barred partitions, can be erected between them, as at the Royal Mews. Jordanus Rufus (13th century author of a horse treatise) describes that colts in training were stabled, hobbled, and tied to the manger on straw bedding. Another agricultural treatise by Pietro de Crescenzi (13th-14th century author) says there should be wooden partitions between horses to decrease fighting when stalled. Some of the incisors (front teeth) of horses from several of the castles covered by the Warhorse project’s zooarchaeological work show damage from regularly eating from a hard manger, cribbing (crib-biting), fighting between bars on adjacent stalls, or maybe even from repeatedly rubbing bars out of boredom (Fig. 7).

Hay rack and manger at Dunster Castle with gnawing damage

Fig. 7. Hay rack and feed manger at Dunster Castle, with some gnawing damage visible along the manger’s edge. Photo by Oliver Creighton.

Damaged lower incisors from Elverton, Westminster. Enamel is worn smooth on the front of the teeth

Fig. 8. Damaged lower incisors from Elverton, Westminster. Enamel is worn smooth on the front of the teeth. Photo by Kate Kanne.

Box stalls, or loose boxes, on the other hand, take double or more of the width of standing stalls, and are less efficient to muck and more costly in terms of bedding. The norm for many horses now, they allow more room for each horse to move and rest, including escape from difficult neighbours. They are better for foaling mares and mare-foal pairs, individualised veterinary care, and for horses that have limited turnout (time loose out of doors), though daily turnout is always best. Tie stalls are used now for horses that are brought in to feed and work each day, with ample turnout otherwise. Medieval horses were used daily for everything from transport to war, and had to be ready to go at a moment’s notice, especially those of the nobility. No one likes having to catch horses in a hurry in a field, as they may not feel the same pressure to comply with your demands!

The earliest historical record describes Kenilworth’s stables as 180 ft long by 21 ft wide in 1563, with 30 ‘rooms’ for great horses, and 20 rooms for geldings (Chirk Survey 1563). Given the size of the building as it stands today, about 49 m long by 10 m wide, these numbers can be accommodated in a number of configurations, but likely with mostly tie stalls, on both long walls, a few box stalls, and a wide aisle between (6 ft minimum is the modern standard, 10 ft is common – the one at Kenilworth could have been up to 12 ft wide). A hayloft and quarters on a second storey were also likely, with the team spotting what could have been used to drop down hay from the hay loft (Fig. 9). 

Possible hay drop at Kenilworth Stables

Fig. 9. Possible hay drop at Kenilworth Stables, Photo by Oliver Creighton.

A smaller, earlier 14/15th-century stable building was found under Kenilworth stables during excavations in 1970-84 (Howard et al. 2006). Robert Dudley shortened his father’s original stable, probably by one bay on the north end. During later excavations for the visitors’ centre at Kenilworth, evidence of the earlier building in the same area was also found, with the later medieval building similar in ground plan to the current standing structure, including a paved floor with drainage (Ellis 1995: 114). Based on this, Ellis concluded that the horses were kept in stalls facing the curtain wall and suggested that they would have potentially been 2 m wide to accommodate 25 stalls within the stable.  

At Dunster, much of the original configuration is still standing, with the portion that is occupied by the gift shop added later. It has a single row of tie stalls facing the wall with hay racks and mangers in place and a good visible drain (Fig. 10). There are different configurations of stone from where the horses stood to the aisle to increase drainage, and horse comfort. The floor at Kenilworth is currently stone, but a line of demarcation around the base of the interior may indicate that plank wood flooring was used there, which is known for other stables for royal horses.

Tie stalls with drain and cobblestones at Dunster Castle

Fig. 10. Tie stalls with drain and cobblestones at Dunster Castle. Photo by Oliver Creighton.

Historical accounts from the 14th-century list the number of horses that medieval castles could house, from six to 779(!) during battles (Colvin et al. 1963; Roberts 2002). Temporary stabling, or simply tying horses to quickly erected hitching posts, would be necessary for castles in the active throws of battle or siege. Kenilworth is the largest known stables, but comparable to those at other high-status sites of the 16th century, like those at New Hall, Essex, and at Reading, Berkshire (Colvin et al. 1982: 173, 221). 

Barnard Castle, Durham, with Saxon-17th century occupation, has had 6 different stables identified by archaeological and historical records, including 2 buildings in the Inner Ward and four in the Outer Ward (Little, Long, Palfrey, and Great). This corresponds well to the large number of horse bones reported at Barnard, especially in the later period (Austin 2007:590). Barnard Castle was also home to the royal herd of the Beauchamp stud. In 1325-1326, according to the records of the keepers of horse, and not including riding animals of the castle residents, the herd comprised two stallions, 35 mares and 50 colts and fillies under the age of two (Austin 2007:105-6). The multiple stables could suggest that the value, type, or utility of said horses would dictate where, or if, they were stabled, as breeding and youngstock were generally kept in the parks (on pasture). Potentially, there were different stables for the castle residents, visitors, breeding stallions, and/or foaling mares of the royal herd, or it may be that additional stables were added through the life of the castle as horse breeding increased at Barnard through time.  

What have we learned about medieval stables?

Medieval stables were well and purposefully built, matching and exceeding modern requirements of stall space per horse and aisle width. They appear to be designed for the working horses of the castles, and for those owned by visiting guests. Though generally built for the convenience of people, medieval stables also account for the comfort of horses. The builders of the stables at Kenilworth and Dunster were experienced horsemen; they kept in mind that the equine inhabitants were very valuable and necessary for daily life of medieval castles, and designed the stables as such.


Austin, D. 2007. Acts of Perception: A Study of Barnard Castle in Teesdale, Volume 1 and 2. Durham: The Architectural and Archaeological Society of Durham and Northumberland.

Brady, C. A., K. S. Kanne, and M. A. Russell. 2002a. Introduction to Housing for Horses. Purdue University Cooperative Extension Service.

—–. 2002b. Introduction to Horse Management. Purdue University Cooperative Extension Service.

Chirk Survey. 1563. Chirk Castle Ms F13310. National Library of Wales.

Colvin, H. M., Allen Brown, R., and A. J. Taylor. 1963. The History of the King’s Works, Volume 1 and 2: The Middle Ages. London: Her Majesty’s Stationary Office.

Ellis, P. 1995. The Elizabethan Gardens and Leicester’s Stables at Kenilworth Castle, Transactions of the Birmingham and Warwickshire Archaeological Society 99:81-116.

Fehr, R. L., Walker, J. N., Duncan, G. A., and W.E. Wise. 1993. Housing for Pleasure Horses. Lexington, KY: University of Kentucky Cooperative Extension Services.

Howard, R. E., C. D. Litton, and A. J. Arnold. 2006. Tree-Ring Analysis of Timbers from Lord Leicester’s Stable, Kenilworth Castle, Warwickshire. Research Department Report Series 21. London: English Heritage.