The north side of a well-preserved west-east Roman drain F1, constructed of brick set in opus signinum mortar, was uncovered during rebuilding work at the rear of 99 High Street, Colchester. The archaeological remains were uncovered in modern foundation trench, 12.1m long x 0.62m wide x 1.5-1.8m deep and aligned E to W, at the rear of 99 High Street. The drain remained undamaged during the rebuilding work.<1> The Roman drain was discovered for a distance of 5.46m in the eastern half of the modern foundation trench, and for 1m in the service trench to the east. The total length over which it was traced was 7.4m. It lay approximately 0.9m below the level of the existing yard surface and was 0.4-0.45m wide and was constructed of Roman brick set in a hard pinkish mortar (opus signinum). The drain was encountered during the 1964 excavation, and the evidence from that investigation suggested the drain was in use from the first half of the 3rd century to the mid 4th century. <2> The drain was sealed by post-Roman topsoil (‘dark earth’), 0.6-0.7m thick (Sx 1-Sx 3), which contained a moderate amount of septaria, Roman brick/tile and mortar fragments and gravel. At the western end of the trench, the ‘dark earth’ was cut by the 1964 excavation trench, which was 1.85m wide (Sx 2). The ‘dark earth’ was sealed by modern rubble make-up and the existing concrete yard surface, 0.2-0.5m thick. Other Roman deposits were observed in the sides of the modern foundation trench, although it was not possible to examine these in any detail. There were also patches of reddish-brown Boudican destruction debris at a depth of approximately 1m below the modern ground level. A well-preserved patch at the western end of the trench was sampled and fragments of burnt daub, plaster and Roman brick/tile, as well as a small quantity of animal bone and a piece of human bone, were recovered. The burnt structural daub and other building materials probably derive from wooden buildings which had existed nearby. Elsewhere in the trench, the Boudican destruction debris was more mixed, with smaller inclusions, and had probably been re-worked to form a make-up layer. At the western end of the trench, the well-preserved patch was sealed by a clayey make-up layer 0.15m thick with small fragments of burnt daub, which clearly included some re-worked Boudican destruction debris. Much of the remaining Roman deposits in the sides of the trench, both pre- and post-Boudican, consisted of thick gravelly layers (Sx 3). These did not have well-defined metalled surfaces, and so could have been make-up rather than street metalling. A total of 19.3g of bone remains was received for analysis. One (15g) is a fragment of right human ilium from an adolescent (15-20 years) to adult (>20 years) individual of uncertain sex. No morphological or metric criteria were available to determine gender. The rest of the fragments (4.3g) were cattle remains.