Evidence of substantial Roman occupation at S. Lorenzo had already been revealed by the surveys undertaken in 2005 which produced a large quantity of Imperial and late antique pottery. The data from trenches V and VI confirmed that the site was occupied between the 2nd-6th century A.D. the discovery of a tank in _opus latericius_, lined with _opus signinum_ and probably communicating with other features, suggests the existence of an organised structure for production activities, perhaps linked to the processing of agricultural products. This structure may have been part of the _pars rustica_ of a villa (attested by mosaic _tesserae_ and precious marbles), or a _vicus_ like settlement, or a larger type of settlement. Matching data from the survey and the excavation attests the presence of a vast settlement. The problem was confronted regarding the establishment of a connection between the archaeological remains that had come to light and the toponym _Carmeianum_, always attested in the sources as an adjective describing the territorial reality ( _ager_, _saltus_, _ecclesia_), rather than a specific settlement, but however evoking a reality of the Daunian landscape. Furthermore a widespread topographical distribution of Roman and late antique settlements in the area is suggested by the discovery in trench V of a second stratigraphic sequence, perhaps of Roman or late antique-medieval date. Although it was clear that the medieval remains (11th-12th century?) obliterated the Roman and late antique ones, it was still difficult to reconstruct a precise chronology. In trench VI three underground structures emerged cut into the calcareous bedrock. These were perhaps pits-silos for storing and preserving cereals, or cisterns, or were however structures linked to the provision of food or water for the settlement.