Salamanca lies on the right bank of the river Tormes, a tributary of the Douro, on the northern sub-plateau of the Iberian peninsula (fig. 1). Although hardly mentioned in Roman historical sources, it is a reference point for work on Roman territory because the surveyor Frontinus (De Agrorum Qualitate [ed. Thulin 1971] 1–2) used Salmantica (in Lusitania) and Palantia (in Citerior) to exemplify ager per extremitatem mensura comprehensus, the system of land organization characteristic of stipendiary cities. Frontinus was writing in Flavian times, but the creation of ager mensura comprehensus in Lusitania occurred in the Augustan period, as is confirmed by remarkable epigraphic documentation. In N Lusitania, a total of 11 boundary-stones (termini Augustales) are known, nine from the reign of Augustus (and two of these provide explicit reference to Salmantica) and two from that of Claudius. The dates provided by Augustus’ tribunicia potestas allow us to date the surveying operations delimiting the urban territories to between A.D. 4–5 (the inscriptions from Peroviseu and Ul) and A.D. 5–6 (the inscriptions from Sao Salvador, Ledesma, Ciudad Rodrigo, and the new one from Jarandilla de la Vera). The Augustan ager mensura comprehensus may have conditioned the model of the subsequent rural settlement by creating a framework for territorial occupation being organized around the villa from the Flavian period on. The villa would dominate the rural countryside, until it disappeared around the first decades of the 5th c. as part of a process that can be associated with the breaking down of imperial authority and the arrival of the Germanic peoples in the year 409. Almost nothing is known about Salamanca’s territory during the Islamic occupation until the first official repopulation took place under the king of Leon, Ramiro II, in 939–49. The lack of attested settlements in the Douro valley between the 8th and 10th c. is a key question for the organization of the border area between the Islamic state of Al-Andalus and the kingdom of Leon, but scholars generally reject the thesis formulated in 1966 by C. Sánchez Albornoz, which tended to present the lands of the Douro valley as practically depopulated.