Evaluation by T Collie of SCAU. The results show prehistoric and/or medieval or early post-medieval activity across the area. Most of the features were identified in trenches located across the western half of the southern field, and they include a number of substantial ditches of probable medieval or early post-medieval date.

Evaluation by W Weller of SCAU revealed no archaeological features. Some finds were recovered from the subsoil and spoil heaps including a small amount of struck flint and post-medieval/modern ceramics.

Historic building recording by S Angell of CBAS. The extant buildings appear to date from the Second World War and follow a fairly standard ‘ministry building’ form, although most have been considerably modified. Many associated buildings are known to have been demolished as the function of the site changed.

Evaluation by P Capps of WA revealed no archaeological features or deposits. Anomalies identified in a previous geophysical survey were revealed to be modern land drains and their associated cuts. The evaluation did reveal a redeposited clay layer relating to the construction of the 19th century railway embankment that runs along the south-eastern edge of the site.

Evaluation by T Vieira of ASE on land to the west of the lane revealed the base of a small, rectangular, 18th century, brick structure, despite considerable horizontal truncation. Its function could not be determined although its date, together with cartographic evidence, suggests that it is associated with Worcester Park gunpowder mill. No other finds or features were recorded.

Evaluation by S Porter of MOLA recorded a rough deposit of chalk considered to represent the foundation of a structure with an associated cobblestone surface. The orientation and location of the possible structure was similar to that of a large building marked on Joseph Hodgkinson’s 1781 map of the manors of Molesey Matham and Molesey Prior. To the south, several shallow, linear bedding trenches were recorded. No dating evidence was recovered from any of the features.

A phase of post-demolition evaluation, test pits and strip, map and record excavations by W Perkins of PCA, following pre-demolition evaluation in 2016 (SyAC 101, 214), showed severe disturbance across much of the site owing to its former industrial uses. However, initial evaluation showed that areas of undisturbed land were present and the later phase of work revealed more of a series of natural palaeochannels, first recorded in 2016, threading across the north-west of the site.

Evaluation by K Bower of PCA revealed extensive modern truncation owing to gravel extraction and landfill across the north and west of the site. A small quantity of residual Iron Age/Early Roman pottery was recovered from a pair of parallel ditches, shown to correspond with a former field boundary on the 1806 Walton Enclosure Map.

Evaluation by G Seddon of PCA recorded no finds or features of archaeological interest.

Community test pitting in 2016 and 2017, directed by A Sassin and D Graham of SyAS. A total of fourteen 1m2 test pits excavated at Farnham Park, 5 Castle Street, the Windsor Almshouses, St Andrew’s Rectory on Upper Church Lane, the Old Vicarage and Coxbridge Farm revealed post-medieval and modern disturbance at the majority of the locations. Post-18th century finds were recovered from all locations but probable 12th–13th century contexts were recorded at the three sites within the known medieval town core (St Andrew’s Rectory, the Windsor Almshouses and 5 Castle Street).