The coatis are social carnivores of the family Procyonidae that inhabit the Neotropics (from Arizona, U.S., to northern Argentina and Uruguay). Traditionally, three species of coatis are placed in two different genera: Nasua (Nasua nasua distributed in South America; Nasua narica, distributed in Central America) and Nasuella (Nasuella olivacea, distributed in the Andean Cordilleras of Venezuela, Colombia, and Ecuador). Recently, a supposed new species of Nasuella has been reported (N. meridensis) in the Venezuelan Andean Cordillera. However, the systematics and the evolutionary history of the coatis are extremely confusing. Herein, we present an phylogenetic analysis of 345 coati specimens sampled from southern Mexico to Uruguay sequenced at eight mitochondrial genes (ND5, ND4,Cytb, D-loop, COI, COII, ATP6, and 12s rRNA). We detected 19 main haplogroups (10 haplogroups in N. nasua, five haplogroups in N. olivacea, and four haplogroups in N. narica) and eight to ten sub-haplogroups (five-seven in N. nasua, and three in N. narica). The oldest and original haplotypes were from Colombian and Ecuadorian Andean Cordillera N. nasua specimens, followed by some haplotypes of specimens with the phenotype of N. olivacea also in the Colombian and Ecuadorian Andes, but with DNA closer to that of N. nasua. Thus, the Northern Andes seems to be the original point where the mitochondrial DNA coati diversification began around 13 millions of years ago (Late Miocene). This is in contrast to the traditional paleontological view, which considers that coati appeared in North America and later migrated into South America during the Pleistocene. The present molecular results more strongly support a single unique genus (Nasua) rather than the traditional two genera. If we apply the Phylogenetic Species Concept (PSC), at least 19 species of coatis should be described. However, the evolutionary history of the coatis is complex and as such, we consider that a greater preponderance of evidence supports one to three species. Two (or three) N. olivacea haplogroups were more related to some N. nasua haplogroups than to other N. olivacea haplogroups. This means that N. olivacea is polyphyletic or there were some Andean N. nasua groups that have evolutionarily converged to similar phenotypes as those of the “true” N. olivacea. In fact, we detected the presence of a “true” N. olivacea haplogroup distributed within the Cuzco Department in southern Peru. Thus, this is the first study to molecularly detect this taxon in Peru. Furthermore, we detected N. nasua specimens with haplotypes very related with those of the original N. narica haplogroup. The first N. narica haplogroup was detected in the trans-Andean and Pacific Ecuador. This haplogroup later generated the Central American N. narica. In Central America, we clearly detected three haplogroups, two of them closely related (northern Costa Rica, Nicaragua, El Salvador, Honduras, Guatemala, and southern Mexico). However, the fourth N. narica haplogroup present in southern Costa Rica, Panama, and northern Colombia was mitochondrially introgressed by N. olivacea. Therefore, N. narica originated in South-America and later migrated to Central America.