An archaeological watching brief was undertaken by Colchester Archaeological Trust in the Upper Castle Park, Colchester, Essex between April and June 2011. Seventeen monitoring visits were made during landscaping works associated with the laying out of the new play area and the park service yard, including the installation of a new access road, metal and concrete fences, hardstanding and services.<1> Archaeological remains dating to the Roman period were encountered within the area stripped for the new access road and in the base of two of the post-holes. Elsewhere, ground disturbance was confined to the dark grey/brown sandy-silt which overlies the archaeology and is interpreted as a modern topsoil with an underlying post-Roman dark earth. Between 0.20m and 0.35m of topsoil was removed for the construction of a 3.5m-wide and 125m-long access road to the park service yard. The removal of topsoil for the new access road in the southern half of the former putting green revealed archaeological remains at a depth of 0.3m below modern ground-level. In this area, a lighter soil was exposed and, within this lighter soil, a small area of tessellated floor roughly 0.4m² was exposed. This Roman tessellated floor had been uncovered during excavations undertaken in 1927-9 (Hull 1958, 91; <2>). The remains of a probable Roman clay-block wall were also identified just to the south of the tessellated floor. The depth of the tessellated pavement bgl was significantly shallower than it had been in 1927-9 when the floor was at depth of ‘18 inches’ (0.45m; Hull 1958, 91; <2>). This change suggests that the ground-level has been reduced in this part of the park, presumably when the putting green was made. A geotextile material was placed over the tessellated floor and the remains of the probable clay-block wall before the road material was laid over the top. An area of approximately 450m² was reduced by 0.25m so that tarmac and gravel could be laid to create new areas of hardstanding and car-parking in the park service yard (otherwise known as the park ‘nursery’). The excavated material was mostly modern topsoil and old surfacing materials such as gravel. A small area of light brown/yellow clayey-silt containing frequent charcoal flecks, Roman brick and a sherd of Roman grey ware pottery was uncovered in the south-western corner of the stripped area. However, this deposit was located just beneath the turf and also contained a tarmac chipping. The deposit was not in situ and was probably disturbed from deeper down during previous works in the park service yard. No other archaeological deposits or features were observed during the monitoring of these groundworks and nor were any other finds recovered. The installation of metal and wooden panel fencing between the new play area and the park service yard involved the monitoring of the excavation of post-holes for the new fence posts. The excavation of eighteen post-holes for metal fence posts and six post-holes for concrete fence posts was observed during five archaeological monitoring visits. The post-holes for the metal posts were hand-excavated to a depth of between 0.60m and 0.70m and were approximately 0.25m in diameter. The post-holes were all excavated through the dark greyish-brown sandy-silt topsoil which contained rare fragments of brick and tile (mostly Roman but some post-Roman), animal bone and modern pottery and glass. The datable finds recovered during the monitoring of post-holes 1-7 were all Roman and included CBM, pottery fragments, animal bone, oyster shell and mortar. A light brown/yellow sandy-clay with frequent oyster shell fragments was encountered at the bottom of post-holes 6 and 7. A similar deposit had been noted in test-pits TP3, TP5 and TP6 excavated in February 2011 (CAT Report 575; <3>). The deposit is most likely to be Roman. The six post-holes for concrete posts near to the new entrance to the park service yard were of the same depth but slightly larger in diameter (approximately 0.4m) than the post-holes for the metal fence posts. The post-holes for the concrete posts were excavated through the same topsoil and at least three of them were located within the backfill of a new service trench. No archaeologically significant deposits were encountered but a few Roman finds, including a pottery sherd from a relatively rare import type (from Central Gaul in blackslipped ware) at Colchester, were recovered (finds no 9). A trench was dug to construct a new kerb along the eastern edge of the existing access road into the park service yard. The trench was 100m long and 0.2m deep. It was monitored on a number of occasions, but it was not deep enough to reach any archaeological deposits below the topsoil. A shorter but substantially deeper trench was excavated from an existing foul-water pipe inspection chamber to the northern edge of the new play area in advance of the installation of a temporary toilet block. The trench was 15.0m long and roughly 0.5m wide. It varied in depth from 0.4m below modern ground-level in the south-east to roughly 0.8m in the north-west near to the inspection chamber. No archaeological deposits were observed in this trench. However, the conditions during the inspection of the trench were not ideal as dust, loose stones and areas of trench edge collapse had covered most of the base of the trench before the trench could be examined archaeologically.