Bats number more than 1,500 species, over a fifth of all living mammals. They are the only mammals capable of true powered flight, and almost all of them find their way by echolocation, emitting sounds and listening for the echo. Where they come from has been an open question for decades: North America, Africa and Asia have each been proposed.
The problem is the fossils. The oldest date to the early Eocene, between 56 and 52 million years ago, and come from Asia, North America, Europe and Australia. They are already complete bats, able to fly. Before them the record is almost empty, made of teeth and fragments.
The Bat1K consortium, 137 researchers led among others by Emma Teeling of University College Dublin, published the first phase of its work in Nature on 23 September 2026. It comprises 103 chromosome-level genomes, 42 of them new, covering all 21 recognised bat families. To these the group added a matrix of 699 anatomical characters for 65 species, 44 of them fossils.
The method matters as much as the result. The authors placed fossils and living species in the same tree, with a model that combines the birth of species, their extinction and the probability of leaving fossils. On top of that tree they ran a model of dispersal between continents that accounts for continental drift and for the mobility of animals that fly.
The result: the common ancestor of bats lived in Europe in the late Palaeocene, with a posterior probability of 99.2 per cent. Its descendants most probably moved into Africa. From that European and African hub, several independent expansions reached the Americas, Asia and Australia in a narrow interval at the start of the Eocene. Yinpterochiroptera, the suborder that includes flying foxes and horseshoe bats, arose in Africa about 57.1 million years ago. The model rejects a direct crossing from Africa to South America over the Atlantic.
Then there is Vielasia, a fossil bat about 50 million years old that could echolocate. The analysis places it on the oldest branch of the tree. If the placement holds, echolocation is older than the diversification of modern bats. The support, however, is moderate: 68 per cent.

Now the date. The paper says «late Palaeocene» in words, and the node ages sit in the figures. The university press releases say «around 65 million years ago» or «65 to 60». On the International Chronostratigraphic Chart the late Palaeocene runs from 59.2 to 56.0 million years ago; 65 million years ago is a million years after the dinosaurs died out, in the early Palaeocene. The two wordings do not match, and the difference should be cleared up with the authors before a figure is printed.
The genome also tells a messier story than expected. For the Myzopodidae, the sucker-footed bats of Madagascar, only 22 per cent of gene-by-gene trees support the proposed placement, though it is still the best supported. On the relationships between the major superfamilies, different chromosomes tell different stories. The authors put this down to ancient interbreeding between lineages and rely on a region of the X chromosome where recombination is nearly absent and the true history of species is better preserved.
One question of method remains that the authors do not address in the text we read. The European Eocene fossil record is unusually rich, and a continent that has been dug and described more carries more weight in any model that uses fossils. The model accounts for geography and flight; how far it accounts for uneven collecting effort is a question worth asking.

The work also leaves a resource: genomes, an ancestral karyotype reconstructed as 26 chromosomes, public code. It will be used to study what sets bats apart among mammals: longevity, up to eight to ten times what their size would predict, and tolerance of many viruses.
The where seems settled. The when, for now, is written two ways.

References
Morales A.E., Liang Y., Thomas W.R. et al. (137 authors; senior authors Ray D.A., Vernes S.C., Dávalos L.M., Hiller M., Teeling E.C.), «Reference genomes and fossils revise bat family phylogeny and biogeography», Nature, 658, 8134, 141–152, published online 23 September 2026, open access. doi: 10.1038/s41586-026-11007-3
Hand S.J., Maugoust J., Beck R.M.D., Orliac M.J., «A 50-million-year-old, three-dimensionally preserved bat skull supports an early origin for modern echolocation», Current Biology, 33, 4624–4640.e21, 2023 (cited in the paper; not read directly).
University College Dublin, «Bats originated and took first flight in Europe, new UCD-led research suggests», press release, September 2026; republished by ScienceDaily on 30 September 2026.