ACO is a multy-period field survey carried out by the Universities of Amiens, Cambridge and Oxford in the valley of the Somme. This interim report summarises the aims and research design of the project, and reports on the seasons held in September 1991 and September 1992. The aims are to assess how complete and how representative a record of the ancient landscape is provided by aereal photography ; to investigate the transformation of the late prehistoric landscape into the Roman one, and to date the end of this classical landscape ; to investigate the links between the transformation of rural space and the growth and decline of Samarobriva. The method is to survey four designated transects, each one kilometre wide and ten in length, running north-south, at equidistant distances (ten kilometres) from Amiens, giving a sample of the landscape and of the structures (native farms, villae) known by R. Agache records. The fields are surveyed on a twenty metres grid, the aim being to assess variations in the density of artefacts distributions, rather than to identify and to delineate sites as such. After the identification and datation of the finds, a computerised archive is created, which will provide the basis of the final report. It is difficult to draw conclusions after only two seasons, where just over 300 ha. have been surveyed, in a project planned to last over five seasons. It seems that the number of previously unknown sites is very small ; there are indications that some native farms remained occupied well into the Ilnd century at least, some villae into the IVth century and possibly beyond ; it seems that the dispersed settlement pattern continued to structure the landscape, in spite of the development of urbanisation. But we must stress the provisional nature of these findings.