This volume is a very welcome addition to the literature on the villas and related sites of Roman Britain. As the only substantial villa in the care of the National Trust and open to the public, Chedworth has paradoxically been ill served by the lack of a substantial report and survey of the excavations and surviving remains. This is put right by this monograph. Many old excavations and smaller interventions have been brought together, and there is a substantial conspectus of the finds, both those in the site museum from the original uncovering of the remains, and from the more systematic recent excavations. Not everything is in this volume, however, as the editors themselves point out. The result is substantial; in fact, rather a weighty and awkward-to-handle tome, with very long chapters. Two volumes would have been a good idea.

After an introductory survey of the history of investigations at Chedworth, a revealing analysis of the walling elevations by Jason Wood demonstrates how much restoration and reconstruction has taken place since the site was first opened for public display. This chapter is somewhat over-long at nearly 130 pages and has a lot of detailed description better suited to a technical (online) report. Only a summary is really needed here. There are pull-out wall sections and elevations at 1:100 scale, and it would have been good to have equally detailed plans. Chapter 3, covering the structural remains and the evidence for phasing and sequencing, is the heart of the volume, with four phases, in contrast to the well-known ‘fossilised’ plan of the site in the fourth century that figures in previous literature. Dating the phases is not easy, due to the lack of stratigraphic recording in the nineteenth century. It is therefore welcome to have two separate excavation reports at the end of this chapter, by Ron Shoesmith and Maureen Carroll. The latter report is particularly useful in yielding chronological sequences, and also valuable evidence for the features in the courtyards.

Chapter 4, at over 225 pages, is a colossus. It has many good artefactual reports, not least the mosaics (by Stephen Cosh), sculpture (by Martin Henig) and architectural stonework (by Kevin Hayward). There are full and very useful considerations of smaller artefacts, pottery, glass and building materials. However, there are also separate Wessex Archaeology reports on some classes of material, leading to quite a confusing structure. For instance, there are two animal bone reports, which have somewhat different basic counts and statistics, probably because they are from different areas of the villa. We are also referred to a third animal bone report, by Andy Hammon (2002 unpublished report), which is frequently used in the discussion of the bones by Claire Ingrem. Why is it not in this volume, too?

The next two chapters are on the nature and interpretation of the site. Chapter 5, on the development of the villa, discusses the plan, its separate zones, issues of access and usage, to good effect. There is an interesting discussion of the Nymphaeum, including comparison with similar buildings in Portugal and elsewhere, and consideration of the evidence for Diana and Christianity from the site. There is also a valuable drawing together of the evidence for the nature of the site in the later fourth and fifth centuries. This is conventionally when villas are seen to be in decline, with mosaics abandoned or used as mundane working surfaces. This may be the case at Chedworth, as Carroll's excavation suggested, but the recent excavations by Martin Papworth have indicated new construction at the site, possibly in the fifth century. The very late Roman silver ring of a rare type from a metal-detector find near the site (Britannia 45, 2014, 427–8) indicates that Chedworth may have had a relatively wealthy presence at the end of Roman Britain and later.

Chapter 6 is the best in the volume, and Simon Esmonde Cleary's primary authorship of it is evident. He builds on the discussion in chapter 5 to make a good case for a ‘landscape of prestige’, and a ‘landscape of the chase’, and places emphasis on late Roman aristocratic behaviour, drawing extensively on historical sources from Gaul and elsewhere. Chedworth is thus firmly interpreted as a villa ‘amongst the grandest’ both in the region and more widely in late Roman Britain. It is compared with other large Romano-British villas, in a discussion that would have benefitted from some comparative plans.

Chapter 7 is on the historiography of the site's reconstruction and interpretation. This is where Graham Webster's contentious interpretation of the site as a Tempelbezirk is discussed (pp. 533–4). Rightly, the authors reject a Tempelbezirk per se, as it is plainly not like the sites in Roman Germany that have this appellation. The possibility of the site as some sort of hostel for pilgrims, or at least containing hostel-like accommodation within its plan, is not so easily dismissed, however. This gets us to the heart of any discussion of Roman villas and their nature. What, exactly, were they? The Chedworth monograph plumps, as discussed above, for an aristocratic residence with an emphasis, amongst other things, on hunting, similar to La Olmeda, Spain, and many other large villas. But, as the Webster intervention highlighted, there are other possibilities, too, such as a primarily religious aspect (for which a good case can be made at Chedworth; also at Great Witcombe and elsewhere), or perhaps an industrial focus (no evidence for this at Chedworth, but certainly seen at other villa sites). Villas may have had a variety of sources of income and livelihood, much as many country estates do in today's diversified world.

The outer courtyard is the area of the site most relevant to interpretations suggesting a potential hostel for pilgrims, or indeed hunting parties. The volume contains no real discussion of the south range of the outer courtyard, despite it being in nearly all the reconstructions figured in chapter 7. Did it exist or didn't it? A geophysical survey of the outer courtyard, shown in a smallish interpretative plan on p. 195, is derived from an unpublished 1995 University of Birmingham Field Archaeology Report, but that is not included here. Also relevant to the main Chedworth site are the outlying buildings, shown as being intervisible on the GIS viewsheds in chapter 6. The outlying buildings form an important basis for the Walters and Rider ‘religious centre’ interpretation of the site, but since they are outside the National Trust area, do not form a focus of discussion.

Clearly this is not the last word on this fascinating site, and we know that the recent excavations by Martin Papworth will form another publication. The excavations by Roger Goodburn were also not included, and need a proper excavation report, though the brief reports from the annual summaries in Britannia are used in chapter 3. The Villas, Sanctuaries and Settlement in the Romano-British Countryside volume (Archaeopress, Oxford, 2022) is important, too, for Chedworth, with three chapters, by Beeson, Cosh, and Walters and Rider, being directly relevant to the site, offering alternative readings of its interpretation. It is to be hoped that the full range of interpretations of Chedworth will be reflected in the National Trust site literature.

This volume is a fine achievement, and the authors are to be congratulated on bringing a complicated and diverse project to fruition. We now have a very good basis for further analysis of the site.