The landslide of Salo originated from the slump of the southern slope of Mt. S. Bartolomeo (568 m), after the deglaciation of Lake Garda, before 7100 ± 150 uncal yr BP (Beta-34291). The main scarp and the crawn are well developed on the western side and dissected on the eastern. A talus slope develops at the foot of the main scarp. Two different portions of the landslide can be distinguished: the first from about 80 m to 230 m a.s.l., the second below 80 m a.s.l. (the mean level of Lake Garda lies at about 65 m a.s.l.). The latter is completely urbanized while the former is characterized by a network of artificial terraces (reworked for farming purposes). These terraces on the landslide show a different and more irregular pattern respect to the neighbouring areas. This is due to the irregular surface of the slumped area. A narrow belt of artificial debris discarged in the coastal area separates the southern margin of the lanslide from Lake Garda. The geological structure of the area is the most important favouring cause of the slump, even though Pleistocene glacial erosion can be also taken into account. Neotectonic activity or an earthquake could have caused the down sloping of the southern slope of Mt. S. Bartolomeo. The "Scaglia" formation (marl and limestone; Eocene-Cenomanian) overthrusted during the Late Miocene on Oligocene marls (Late Oligocene Middle Oligocene) occurring at the bottom of the southern slope of Mt. S. Bartolomeo. An angular unconformity separates the strongly folded "Scaglia" and the basal conglomerate (Early Pliocene? - Late Miocene ?) of the formation of Mt. S. Bartolomeo (Late Pliocene-Late Miocene ?). The rocks that outcrop along the southern slope of Mt. S. Bartolomeo ("Scaglia" and conglomerate) slumped on the plastic Oligocene marls, carrying down also the surficial deposit of the area (till and stratified debris, dating to the Late Pleistocene and to the Holocene). In fact, Pleistocene and Early Holocene deposits are not present on the landslide of Salo. On the contrary, they are widespread in the neighbouring areas. Furthermore, the southeastern margin of the landslide borders on a lacustrine delta raised above the present lake level. The exact age of this delta is unknown but, according to other evidences of the Garda area (BARONI, 1986), it can be dated to the Early Late Glacial Holocene. The artificial debris discarged in the coastal area between the 11th century AD and the present is more than 13 metres thick. Shells of Valvata piscinalis and Bithynia tentaculata collected at 14.7/15.4 m in a drill carried out in P.za Zanardelli (tav. 1) yielded an age of 7100 ± 150 uncal BP yr (Beta-34291). The shells have been collected from a sandy silt layer at the top of a deltaic sequence. Since the lake level decreased during the Holocene, the deltaic sequence between 13.7 and 16.2 m demonstrates the subsidence of the area. This series has been interpreted as a deltaic sequence resting on the southern portion of the landslide. The date supplies a minimum limiting age for the down sloping. Thus, also taking into account the "reservoir effect" of the lacustrine water that produces older dates for the carbonates of the freshwater molluscs, the direct superimposition of medieval layers on the dated deltaic sequence shows a stratigraphic gap of some millennia. This attests a remobilization of the southern margin of the landlslide between 7100 ± 150 uncal BP yr (Beta 34291) and 880 ±70 uncal BP yr (GX- 14685). If this is the case, one can suggest that settlements existing in the area fell into the lake. The central portion of the coastal area (on which the historical centre of Salo was built) was interested by subsidence between 16th and 20th century. It slipped toward the lake on the occasion of at least two earthquakes (5.1.1892 and 30.10.1901). The latter (VII-VIII MCS) gave rise to the greatest damages of the history of Salo. The subsidence of the coastal area is testified by the annual mean lake levels registered by the hydrometers of Salo and Peschiera. A subsidence of some centimetres is documented between 1953 and 1985. Portions of the landslide were remobilized during the 19th and 20th centuries. Some of them still seem to be in movement (Misse, Guasto and Rocchetta areas). Several Authors refer of a legend according to which the mythic city of Benacus fell in the lake after an earthquake in 243 (or 245) AD. Evidence of this city was never pointed out and many Authors do not trust the legend. Somebody thinks that the ruins of a roman villa at Toscolano are to be related to this city. The geological outline of the area suggest a down sloping of the coast at Salo before 880 ±70 uncal BP yr, 0031/1225 cal AD; GX-14685). This makes us suppose that the mythical city of Benacus, if really existed, could be hypothetically located on the landslide of Salo.