Lay 386 fossil bodies out across two and a half million years and the shape they make is not the one you were taught. There is no tidy ramp, no steady upward slope of ape giving way to almost-human giving way to us, each generation a little taller than the last. Instead there is a long, gentle drift, and then, somewhere around two to two and a half million years ago, a jump. A growth spurt written into the genus, sudden enough to break the line.

That jump is the headline of a study published this week in the Proceedings of the National Academy of Sciences, and it rewrites one of the most familiar pictures in human evolution. The march-of-progress image, the one on the T-shirts, turns out to be wrong in an interesting way.

For decades the field could not agree on something that sounds almost too basic to be contested: did our ancestors get bigger gradually, or did body size leap at one key moment? Some researchers, squinting at the early relatives, saw a slow creep. Others, looking later in the story, saw a step change. The two camps talked past each other for years, each holding a different corner of the same fossil record and insisting it described the whole thing.

Jacob Gardner at the University of Reading thinks the disagreement was never really a disagreement at all.

“For years, different studies have come to different conclusions about whether our ancestors steadily grew bigger over time or jumped in size at some key point in our Homo ancestors,” he says. “We think that’s because everyone was looking at slightly different pieces of a much bigger puzzle.” Put all the fossils on one table, he and his colleagues argue, weigh the competing ideas against each other rather than one at a time, and account for how the species are actually related, and the contradiction dissolves.

Reading the Bones Together

The method matters here, because it is the method that does the reconciling. Gardner’s team pulled body mass estimates from 386 specimens across 21 different hominin species, the sprawling family that includes us and all our extinct cousins. Then they ran the numbers through Bayesian phylogenetic models, a mouthful that boils down to a single useful trick: the models know that closely related species are not independent data points, that a parent and its descendant share an inheritance, and they adjust for it.

They did not run the analysis once. They ran it across 1,000 versions of the family tree and 1,000 versions of the dataset, a way of admitting up front that the fossil record is patchy and that nobody is entirely sure which bone belongs to which species. Uncertainty, baked in rather than ignored.

What fell out the other side was, in the end, both camps being partly right. Body mass did drift upward across the whole hominin story, by roughly a kilogram every million years, slow enough that you would barely notice it generation to generation. But the real action came later, inside Homo, and specifically in the branches after Homo habilis. Homo ergaster and erectus were the first of our relatives to routinely tip the scales past 60 kilograms, weights that would not look out of place in a modern crowd. Allowing for that one late shift explained about 11 per cent more of the variation in the data than pretending it wasn’t there.

And the timing is the tell. The size jump lines up with a cluster of other changes in later Homo: walking upright more efficiently, eating more meat, ranging across far wider territories in search of food and somewhere liveable. A bigger body helps with all of it. Longer legs cover ground; more mass buffers a varied, unpredictable diet.

“Although body mass generally increased throughout our evolutionary history, the most significant shift occurred later within the genus Homo,” says Thomas Puschel at the University of Oxford, a co-author on the paper. The change, he adds, tracked broader developments in how these ancestors moved across landscapes and worked their environments, pointing to a close relationship between body size and major ecological and behavioural transitions.

The Cousins Who Stayed Small

Here is where the story gets its best characters. Not everyone went along with the trend. Two species sat stubbornly off the line, smaller than they had any business being given when they lived: Homo floresiensis, the so-called hobbit of the Indonesian island of Flores, and Homo naledi, whose bones were hauled out of a South African cave system in such numbers that they hinted at something close to deliberate burial. In the models, both turn up again and again as outliers, refusing to grow.

So the clean upward arrow was never clean. Australopithecus, further back still, averaged around 40 kilograms and stood about the height of a child, and stayed there. Branches of the family wandered off in their own directions and did their own thing, size included.

“The human story is not simply one of constant growth, but also of a major change that happened later, within our own genus, while other branches of the family, including some surprisingly small relatives, went their own way entirely,” says Gardner.

There are caveats, and the team is candid about them. Estimating the weight of something that has been dead for a million years from a fragment of femur is an inexact business, and different labs using different reference skeletons can land on numbers that differ by a quarter or more. Whether the gentle background increase is real or an artefact of those wobbles is, frankly, still a bit open. The late leap, though, held up no matter how the team sliced the data, which is about as close to solid as this field gets.

What it leaves us with is a humbler, stranger family portrait than the one on the T-shirt. Not a relay race toward a finish line marked “us”, but a bush of relatives growing, shrinking and holding steady, with one branch happening to surge late and carry its big body out across the world. The hobbits and the cave-dwellers were not failed drafts. They were simply other answers to the same question, and for a long while, theirs worked fine too.

Source: Gardner et al., Proceedings of the National Academy of Sciences (2026), DOI 10.1073/pnas.2521732123

Frequently Asked Questions

So did human ancestors get bigger steadily or all at once?

Both, but unevenly. The new analysis finds a slow, modest increase in body size across the whole hominin family alongside a much sharper jump later on, within the genus Homo, after Homo habilis. The big leap is the better-supported of the two patterns, which is why earlier studies that looked at only one part of the record kept disagreeing.

Why would a bigger body have helped our ancestors at that point?

The size increase lines up in time with later Homo walking upright more efficiently, eating more meat, and roaming across much larger territories. A larger body makes long-distance travel easier and helps an animal cope with a varied, unpredictable diet, so growing bigger appears to have been bound up with a wider shift in how these ancestors lived.

If everything was trending bigger, why were some species so small?

Not everything was. Homo floresiensis and Homo naledi both stayed notably small and show up in the models as consistent outliers, and the early hominin Australopithecus averaged only about 40 kilograms. The picture is less a single rising line than a branching family in which different lineages went their own way, size included.

How confident can we be in body weights estimated from old fossils?

Less than you might hope, which is part of the point. Estimating mass from fragmentary bones is imprecise, and different methods can disagree by 25 per cent or more, so the researchers ran their analysis across 1,000 versions of the family tree and dataset to account for that uncertainty. The late size jump in Homo survived all of it; the gentler background trend is more sensitive to which numbers you trust.