A watching brief excavation was carried out between 2003-5 during the construction of a small housing development on the site of the demolished Handford House, 1 Queens Road, Colchester (relating to CBC planning application F/COL/02/1330). The main part of the excavation was carried out between February and June 2003 while foundations and service-trenches (0.6m wide) were being dug. Subsequent excavations were carried out as and when soakaways and other trenches were dug, between July 2003 and January 2005. <1> The excavation revealed 68 burials in total, consisting of nine inhumation burials (excluding bone from three further inarticulated inhumations), two busta and 57 other cremation burials (seven of which did not contain cremated bone but have been classified as disturbed cremation burials). One possible pyre-debris deposit without cremated bone was also recorded. The two busta are the first burials of this type to be found in Colchester. The excavation consisted of the foundation trenches and service-trenches which equated to approximately 10% of the 68m x 65m site. It is estimated that an open area excavation of the entire site would have exposed approximately 680 burials. The inhumation burials lay at between 0.63m and 1.1m below ground-level. Two of the inhumations were interred in coffins. Cremation burials were scattered fairly uniformly throughout the site at between 0.3 m and 1m below ground-level. Of the 57 cremation burials, 35 were definitely urned, either in a ceramic or glass vessel or in a wooden jewellery box. At least three of the cremation burials contained pots which had been deliberately broken, post cremation, as did one bustum. In two examples, parts of broken pots had been placed to cover lamps, which is good evidence for the lamps having been lit before the graves were backfilled. Twenty of the cremation burials contained one or more ancillary vessels alongside the urn/box, for example, dishes, flagons and small beakers. These may have held food and drink. Between 11 and 15 cremation burials contained deliberately-deposited pyre debris in the pit fill. The debris consisted of small fragments of cremated bone, charcoal and artefacts such as melted glass phials, burnt pottery, nails from wooden boxes, lamps, coins, jewellery, a bone needle, a bone die and hobnails. This material was burnt with the body on the pyre and deliberately buried with the cremated human remains. One of the urned cremation burials was deposited in a large Dressel 20 amphora. Subsequent excavation of its contents revealed a flagon, the neck of the amphora, the cremation urn, a lamp and a dish. Another cremation burial featured a large but broken Brockley Hill amphora which may have contained the cremation urn. Several cremation and inhumation burials and one bustum produced butchered animal bone and fish bones suggestive of grave goods or the remains of graveside feasting. A Roman rubbish-pit containing butchered animal bones may also represent the remains of graveside feasting. The environmental report shows little or no evidence for the deliberate deposition of plant materials on the pyre as offerings to the deceased. Although wood probably formed the main component of the pyres, subsidiary fuels almost certainly included gorse, bean ‘straw’, broom, bracken, dried grasses and grassland herbs. The limited excavation did not show any particular groupings to the cremation burials except for a general thinning out of burials in the north-western corner. The earliest cremations are mid 1st to 2nd century in date and some are definitely pre-Boudican. The cremation cemetery appears to have been in continuous use until the 3rd or 4th century. The southern side of the site featured areas of metalling which may represent a Roman road aligned east to west. The nature of the excavation made it difficult to ascertain whether it was a continuous cambered road, a hollow way or several discrete gravelled areas. Nine adult inhumations (graves) were excavated, all of which were in the northern part of the site. No two burials were the same; the bodies were buried in different positions and on different alignments, and only two bodies had been definitely placed in coffins. One body (male) had been buried wearing a shale armlet. Another (male) was buried wearing hobnail shoes. Apart from being Roman, there is no conclusive dating evidence from most of the inhumations except for three which contained pottery dating them to some time between the mid 2nd to 4th centuries. Two inhumations did contain 1st- to 2nd-century pottery, but this may derive from earlier cremation burials. The shale armlet may either be Late Iron Age/early Roman or late Roman. Therefore it is not possible to say whether the two burial practices of cremation and inhumation were being carried out concurrently or whether all the inhumations post-date the cremations. Study of the human bone shows that those interred had a normal range of pathologies and injuries and that there was nothing unusual in their stature, mortality rates or ratios of male to female. Study of the teeth showed that their diet was rich in carbohydrates, which is normal for the Roman period. The evidence from the glass vessels, small finds, the iconography on the lamps and coins, and the methods of cremation suggests that, in its earliest phases, the burial ground was closely associated with the inhabitants of the Roman colony with their wholly Romanised life-style, rather than the Romano-British (native) population living in the surrounding area. Busta are often associated with military centres and it is a burial rite which is likely to have been brought over from the Continent. Burial in the cemetery ceased before the end of the Roman period when the site was used as a source for gravel. There was very little evidence of activity on the site from later periods. In the 19th century, various large trenches up to 1 m deep were dug which removed some of the cremation burials and disturbed several inhumation burials. It is possible that this was the work of antiquarian George Joslin who lived opposite Handford House. This work followed evaluation in 2002 (CAT Report 210). <2>