The finding of archaeological evidence predating 1Ma and a small hominin species (Homo floresiensis) on Flores, Indonesia, has stimulated much research on its origins and ancestry. Here we take a different approach and examine two key questions - 1) how did the ancestors of H.floresiensis reach Flores and 2) what are the possibilities for estimating the likelihood of hominin persistence for over 1 million years on a small island? With regard to the first question, on the basis of the biogeography we conclude that the mammalian, avian, and reptilian fauna on Flores arrived from a number of sources including Java, Sulawesi and Sahul. Many of the terrestrial taxa were able to float or swim (e.g. stegodons, giant tortoises and the Komodo dragon), while the rodents and hominins probably accidentally rafted from Sulawesi, following the prevailing currents. The precise route by which hominins arrived on Flores cannot at present be determined, although a route from South Asia through Indochina, Sulawesi and hence Flores is tentatively supported on the basis of zoogeography. With regards to the second question, we find the archaeological record equivocal. A basic energetics model shows that a greater number of small-bodied hominins could persist on Flores than larger-bodied hominins (whether H.floresiensis is a dwarfed species or a descendent of an early small-bodied ancestor is immaterial here), which may in part explain their apparent long-term success. Yet the frequent tsunamis and volcanic eruptions in the region would certainly have affected all the taxa on the island, and at least one turnover event is recorded, when Stegodon sondaari became extinct. The question of the likelihood of persistence may be unanswerable until we know much more about the biology of H.floresiensis.