The mammalian fauna of Cenozoic South America is known from the scientific literature to be mainly composed of a series of different aggregate waves of migrants, which originally were recognized and characterized by Simpson (1950) as the first to the third "strata." The living mammal fauna of South America still retains members of the three "strata," but during the Cenozoic the orders of endemic ungulates pertaining to the first stratum became extinct. The living South American ungulates are all northern immigrants from the third stratum. The zenith of the South American endemic or native ungulates can be traced back to the Paleogene. In the Casamayoran South American Land Mammal Age (SALMA) (Eocene) they are represented, conservatively, by five "orders"and more than forty genera. During later times in the Cenozoic there was a significant diminution in the taxonomic diversity, though a reduction in abundance of specimens has not been observed. This was a consequence of worldwide environmental changes that affected the biota during the Eocene and Oligocene transitions. During these transitions the decrease of worldwide temperature (Zachos et al. 2001, 2003) led to the modern "Icehouse-World." As a consequence mammalian communities underwent several changes in their lineages. This global climatic change led to several faunal changes, such as the "Grande Coupure"in Europe. In South America the end of the Eocene was associated with several geological events. Among them, sea-level fluctuation and the rise of the Andes mountain range accentuated climatic changes in South America. So during the Early Neogene (fig. 13.1), for example in the Santacruzian SALMA (Early Miocene), endemic ungulates were reduced from five to three orders (Astrapotheria, Litopterna, and Notoungulata) and nearly 50% of the genera, and holdover lineages became specialized and distinctive, almost "stereotyped"within each group. The evolutionary history of the native ungulates, as well as that of many other South American mammals, is based mainly on the Cenozoic fossil record from Argentina, with a considerable latitudinal range of fossiliferous sites. Hence this history is biased, when we consider all of South America, especially due to the scattered and sometimes scarce fossil record of the more tropical areas. In these outcrops scientific research has been not so extensive as it has been in Argentina, in part due to dif- ficulties in the field (rare preservation and limited access to outcrops; e.g., large fluvial basins subjected to extensive inundation). Considering the Venezuelan fossil record, the oldest mammal is supposed to come from Eocene rocks of Trujillo state. It was described by Patterson (1977) as Proticia venezuelensis and interpreted as a representative of the Pyrotheria, a bilophodont order of South American native ungulates. But its stratigraphic provenance and phylogenetic affinities were questioned by Sánchez-Villagra et al. (2000). Proticia venezuelensis is represented by a jaw fragment with p3-m1. The fossil was collected in 1964 by a second person, several years before Patterson visited the supposed locality in 1970. As related by Patterson (1977), the fossil was reported to be found in a spot along the Quebrada de Agua Viva, which is located about 13 km west of the Cerro La Cruz locality. As described by Sánchez-Villagra et al. (2000), in 1995 Rich Kay, Roberto Lozsan, Rick Madden, and Marcelo R. Sánchez-Villagra visited the purported collection spot of Proticia venezuelensis (Patterson 1977), with the help of Don Guillermo, a local resident who accompanied Patterson in 1970 to the locality. They found no traces of any vertebrate fossil in the area and, based on this observation (also experienced by Patterson 1977, 400), suggested that the provenance of P. venezuelensis should be put into question. P. venezuelensis is more likely to have come from the rocks of the Castillo Formation. Its dental anatomy places the phylogenetic position of this taxon in question. A cuspate tooth form (lack of bilophodonty) and the presence of thick enamel make the possibility of sirenian affinities worth testing (Sánchez-Villagra et al. 2000). This would correspond well with the kind of environment in which Sánchez-Villagra et al. (2000) suggested this fossil was found. Another important outcrop in Venezuela is the Urumaco Sequence in Falcón state (Quiróz and Jaramillo, this volume), which represents one of the few glimpses in the Tertiary of the northern areas. Its fossil mammals indicate a probable Huayquerian SALMA (upper Miocene) for this fauna (Linares 2004), although descriptions of the fossils on which this assessment was based have not yet been published, and correlations between tropical areas and high latitudes are problematic. In tropical areas the La Ventan fauna (Middle Miocene) from Colombia is well known as an early counterpart to the Urumaco fauna of Venezuela. There are also some scattered finds of probable Middle Miocene La Ventan age in Acre, Brazil (Cozzuol 2006), and a recently described La Ventan fauna for Peru is very similar to that of Colombia (Antoine et al. 2007). In Argentina, the Chasicoan and Huayquerian SALMA (fig. 13.1) are chronologically near the Urumaco fauna and represent extra-Patagonian faunas. The Colloncuran SALMA is more or less well known and reflects the changing pattern of the mammalian faunas, specially the native ungulates, in southern latitudes of the continent. The poorly known Mayoan SALMA fauna, Middle Miocene? (fig. 13.1) probably will be of special interest in the future for comparison of one of the last Patagonian faunas before the shift of the mammalian record to more northern areas in Argentina. But to date the Mayoan mammals are not understood well enough, nor are other new records of potential Middle Miocene faunas in Patagonia, to serve as useful antecedents to compare with earlier tropical faunas older than Urumaco. Therefore, given these limitations in knowledge of the Middle Miocene, the Colloncuran SALMA in Argentina and the La Ventan SALMA (fig. 13.1) of Venezuela and Colombia allow us to consider with some accuracy the differences between a southern fauna, probably mainly denizens of open areas, and a tropical one, that probably inhabited more forested and closed regions. In this work we examine what we know about the record of the native ungulates representing the three surviving orders in the Middle Miocene: Astrapotheria, Litopterna and Notoungulata, emphasizing the Urumaco Formation ungulates.