Since the time of Darwin (1859) and Wallace (1869), islands have been regarded by scientists as a prime target for scrutinizing the forces that may influence evolution and diversification and important elements in biogeographic studies. This research aims to scrutinize whether and to what extent the composition and structure of past mammal insular faunas and their changes through time may provide sound clues for inferring the paleogeographical evolution of a region. As a case study, I critically analyzed the dynamics shown by the Plio-Pleistocene mammalian fauna of 3 Western Mediterranean insular districts, the Balearic Islands, Sardinia and Sicily, each characterized by its own peculiar paleobiogeographical evolutionary history. The revision of faunas and the critical analysis of the dispersal ability of the ancestors of island settlers have allowed hypothesizing on the time and mode of island colonization. The results obtained confirm that the early isolation of the Balearic Islands from the mainland led to the establishment of an endemic fauna since the pre-Messinian Miocene (?Astaracian European Land Mammal Age, MN7/8), and that Sardinia has definitely been isolated since the Pliocene, although dispersal events led to some faunal turnovers during the Pleistocene. In addition, the results suggest for Sicily a complex, still imperfectly disentangled history of alternate phases of complete separation and sporadic, more or less difficult connections with southern Italy.