This volume is the latest in the line of urban archaeological assessments produced as a consequence of an initiative launched by English Heritage (now Historic England) in 1992. Assessments of important Roman towns that have reached publication include Cirencester (Darvill, Gerrard, Viner, Holbrook, Gutiérrez and Johnson 1994); Verulamium (Niblett 2005); Colchester (Gascoyne, Radford, Crummy and Wise 2013); and now Bath. A number of other projects were started but have not yet reached a conclusion, although Winchester will soon be published. No further assessment projects of this kind are planned. What has become apparent over the last 25 years is just how difficult these assessments have proved to complete. Few individuals possessed the necessary knowledge and experience of the complex archaeology of the historic towns of England, while even fewer had the time to devote to projects of this sort. Bath is seemingly no exception to this general rule and we learn from the preface that this book has had a far from smooth path to publication. Work commenced in 1997, with a first draft ready in 2001 (by La Trobe-Bateman), but that manuscript languished until 2008 when the project was revived by Niblett, someone well versed in the intricacies of urban archaeology. It has taken a further eight years since then for the book to finally appear, so readers should be aware that the text is essentially as written in 2008, with subsequent work not included. One cannot help but wonder why Peter Davenport, who was so evidently qualified for the task, was not invited to undertake the project from the outset, although he was evidently consulted during the project. Or failing him, N. who made a very successful job of the Verulamium assessment. Tasking a relatively inexperienced researcher with this very demanding commission may explain the delays that have occurred during the production process.

Bath is fortunate in that it has benefited from a number of very high-quality, well-contextualised excavation reports from the pens of Barry Cunliffe and Peter Davenport, who have written syntheses of the Roman and medieval evidence respectively. There is little in this volume that could not be found in their writings, at least for the Roman period. Nevertheless, it is a well-written and illustrated account that provides a useful introduction to the archaeology of Bath for individuals or organisations with little previous experience of working in the city, one of the main drivers for the urban-assessment programme. Of course an assessment of this type can only easily consider work that is documented, hence inevitably it concentrates on investigations that have already been analysed and published. But it is the unpublished excavation archives where most new information potentially resides. The former Bath Archaeological Trust did a considerable amount of work in the 1980s along Walcot Street, which overlies the Roman road leading north from the walled area (important sites include Hat and Feather Yard, Nelson Place, Aldridge's and Tramsheds). These excavations revealed significant evidence for extramural roadside activity that dates back to the first century a.d., and this suburb was likely the ‘urban’ rather than ritual focus of Aquae Sulis. Analysis and dissemination of these investigations, among others, will doubtless provide considerable insight into the sequence of development and the functions performed in the Walcot Street suburb, and potentially shed light on the often assumed, but as yet unproven, military origins of Bath. Work in advance of development in 2012 in Bathwick on the opposite bank of the Avon will likewise be instructive in this respect when it is published. One feels that much value could still be obtained from the archives produced by Bath Archaeological Trust, where future attention could usefully be directed.