The Villány Hills in SW Hungary have the richest archive of Pliocene–Quaternary vertebrate faunas in the Pannonian Basin, mostly in karstic cavities. Here we present three new sites that extend the list of Pleistocene vertebrate locations for the area and add information to the evolution history of the region. In the northern part of the Siklós quarry, bone breccia was found coming from fissures in Jurassic or Cretaceous limestones. Its lithofacies and fossil content are similar to those of other well-known Plio-Pleistocene karst infills of the region. As it contained mostly snake vertebrae, its age could not be constrained precisely. In the southern part of the same quarry, two vertical shaft s were discovered, which are unusual in several respects. Th ey formed in a Middle Triassic dolomite succession, a rock type generally not prone to karstification. Th ey might have been created by gravitational deformation of the relatively steep slope, probably at diff erent times. One of them was closed from above and contained fl owstones precipitated during the late Middle Pleistocene, during the late Rissian MIS7 interglacial. The other one was filled from above with loess, rock fragments and remains of large mammals – Equus cf. ferus, Bos primigenius and Coelodonta antiquitatis –, possibly between 140–40 ka, during one of the stadials of the Weichselian or the latest Saalian. The site shows that fossil-bearing cavities could also form in lithologies not favourable for karstification, which then trapped fossils in a similar way karstic cavities do. In contrast with the previous two and with most of the other known vertebrate sites of the Villány Hills, the Palkonya outcrop is not a karst cavity fill but was deposited on the (palaeo)surface. Bison sp., possibly B. schoetensacki remains were found between the Triassic basement and Quaternary slope sediments, and within the latter succession. The Bison bones are probably Middle Pleistocene or late Early Pleistocene, older than ~300 ka. The overlying slope sediments originate from the reworking of various older deposits. Th ey were covered with loess in the Weichselian (~22 ka ago), then again with slope deposits. The abundance of bones in and around the outcrop suggests that this site acted as a fossil trap as well. Bones probably enriched in the sediments during reworking of older deposits. In cold periods, loess deposition decreased (subdued) the relief through infilling the depressions. With 21 figures and 3 tables.