Two trial-trenches were excavated by Colchester Archaeological Trust in August 2012 in advance of the construction of a performing arts block at Colchester Royal Grammar School, Colchester. Both trenches were 1.2m wide and were carefully located between the tree root protection zones present in this area. T1 was 10m long, aligned NE to SW, and was positioned at a right angle to a Roman road known to partly lie beneath the proposed development site. T2 was 6m long, aligned NW to SE, and positioned on the southwestern side of the Roman road. <1> Below the turf, there was a thick layer of dark grey-brown sandy loam topsoil containing modern materials such as plastic and roof slate (L1) (Fig 5). L1 overlay a similar topsoil layer which contained notably fewer inclusions than L1 and no modern materials (L2). The two topsoil layers were present in both trenches and had a combined thickness of over 1m (Fig 5). Beneath L2, a slightly lighter soil containing frequent Roman finds directly overlay Roman contexts (L3) (T1). It was not possible to clean or closely examine large areas of the trenches due to the health and safety implications of their considerable depth. As a result, all recording was undertaken from the top of the trenches. Most of the finds recovered were collected from the upcast soil during the excavation of the trenches with the mechanical excavator. A NW to SE orientated Roman road F1 was uncovered at the NE end of T1, at a depth of between 1.3m and 1.6m below modern ground level. The road was constructed of compacted gravel, covered in a 20mm thick layer of pea grit, and was relatively flat and smooth. Near the south-western edge of the road, the compacted gravel appeared to have been cut by a ditch on a parallel alignment with the road (F2). A significant quantity of Roman finds, including numerous animal bones, was recovered from the upper fill of F2. To the south-west of the Roman road, a light yellow/brown sandy silt containing mortar, oyster shell and charcoal inclusions (L4) was identified at the same depth as the Roman road. This distinctive layer was encountered in both trenches at a similar depth (c 1.3 below modern ground level). Three sondages were excavated into L4 with the mechanical excavator. The sondages in the south-western end of T1 and the north-western end of T2 were excavated to a depth of 2m and 1.8m below modern ground level respectively. L4 was found to continue below these depths in both sondages. The considerable depth of the Roman contexts in this area is attributed to the deposition of additional topsoil on the site within the past 80 years. The Roman road was first identified by A F Hall in the 1930’s, with further excavations taking place on the site of the road in the 1950’s. Two backfilled archaeological trenches from the 1950’s excavations were identified. During the excavation of T2, the top of a backfilled trench was encountered c300mm below modern ground level (F3). F3 had been backfilled with a mixture of topsoil and a light coloured soil containing frequent Roman finds and inclusions comparable to L4. F3 had vertical sides and continued below a depth of 1.5m below modern ground level. To the north-west of F3, a dense concentration of finds, almost all of which were Roman pottery sherds, was encountered (F4), which was presumably discarded into the trench with the backfill. Most of the pottery was located beyond the southern limit of excavation, but two large bags were recovered as a sample. A large quantity of Roman-period finds was recovered during the evaluation. One find of particular interest is a piece of worked fossiliferous limestone, which might be from Purbeck in Dorset and is from the outside of a building or monument, possibly a Roman tomb.