Humans and Neanderthals had sex, but not very often

By Ed Yong | September 12, 2011 3:00 pm

Tens of thousands of years ago, our ancestors spread across the world, having sex with Neanderthals, Denisovans and other groups of ancient humans as they went. Today, our genes testify to these prehistoric liaisons. Last year, when the Neanderthal genome was finally sequenced, it emerged that everyone outside of African can trace 1 and 4 percent of their DNA from Neanderthals.

The discovery was a vindication for some and a surprise to others. For decades, palaeontologists had fought over different visions of the rise of early humans. Some championed the “Out of Africa” model, which says that all of us descend from a small group of ancestors who came out of Africa, swept the world, and replaced every other group of early humans. The most extreme versions of this model said that these groups never had sex, or at least, never bred successfully. The alternative – the multiregional model – envisages these prehistoric groups as part of a single population that met and mated extensively.

To an extent, these are caricatured versions of the two models, and there are subtler variants of each. Still, early evidence seemed to support the extreme Out of Africa version. When scientists sequenced the mitochondrial genome of Neanderthals (a small secondary set of genes set apart from the main pack), they found no evidence that any of these sequences had invaded the modern human genome. The conclusion: Neanderthals and humans never bred.

The full Neanderthal genome disproved that idea, but it also shifted the question from whether humans had sex with Neanderthals to just how much sex they had. As I mentioned in New Scientist earlier this year, modern humans were spreading into areas where Neanderthals existed. “It doesn’t necessarily take a lot of sex for genes from a resident population to infiltrate the genomes of colonisers. When an incoming group mates with an established one, the genes they pick up quickly rise to prominence as their population grows.”

Now, Mathias Currat from the University of Geneva and Laurent Excoffier from the University of Berne have weighed into the debate. They simulated the spread of modern humans from Africa and their encounters with Neanderthals throughout Europe and Asia, to work out the levels of sex that would have transferred Neanderthal genes to modern genomes at their current level.

The duo concluded that sex between the two groups was somewhat of a fringe activity. Fewer than 2 percent of the possible sexual encounters at the time happened between a human and a Neanderthal and produced a fertile, healthy hybrid child. That’s a conservative estimate – the true odds might have been even lower. “Such interbreedings were strongly prevented or very rarely successful,” says Excoffier.

Even if the odds of successful interbreeding were just 5 percent, Neanderthal genes would make up the majority of the human genome today. As it is, a lack of viable sex explains why none of the Neanderthals’ mitochondrial DNA made its way into modern humans, and why so little of their main genome did.

Currat and Excoffier suggest that either modern humans and Neanderthals didn’t have sex very often, or their hybrids weren’t very fit. They favour the first idea. According to their model, it would only have taken between 197 and 430 liaisons between ancient humans and Neanderthals to fill 1-3 percent of modern Eurasian genomes with Neanderthal DNA. Considering that they two groups probably interacted for 10,000 years or so, it would have been enough for one human to sleep with one Neanderthal every 23 to 50 years.

David Reich, one of the leaders of the Neanderthal sequencing effort, is not surprised. “People today mix with people who speak their language and share similar traditions, and these groups were more diverse than modern ones,” he says. “This says to me that where modern humans and Neanderthals were in the same area of the world, Neanderthals would tend to mix with Neanderthals and humans would tend to mix with humans. I would be shocked if they mixed randomly.”

Reich is more interested by the duo’s other conclusion: that Neanderthals and humans continued to mate, albeit rarely, throughout Europe and Asia, all the way from Spain to the Himalayas. This contrasts with Reich’s own work. When he published the Neanderthal genome with Svante Paabo, they suggested that the two groups only bred when humans moved out of Africa and encountered Neanderthals somewhere in the Middle East. This explains why everyone in Europe and Asia has the same level of Neanderthal genes in their DNA – their ancestors picked up those genes only once and spread them to all their descendants.

Currat and Excoffier’s conclusions are clearly different. Paradoxically, they suggest that Neanderthal-human sex was less successful and probably rarer, but also more widespread. There’s an easy way to find out which idea is correct. If Currat and Excoffier have it right, then Europeans and Asians should have picked up different Neanderthal genes. That’s exactly what John Hawks found. In his own analysis of the Neanderthal sequences, he saw that Europeans and Chinese people seem to carry different Neanderthal genes.

This is a fantastic time to be interested in human origins. We went from humans replacing Neanderthals, to humans mating with them and picking up their genes. We learned that this happened with the Denisovans before humans spread throughout the Pacific. We learned that inter-group sex happened in Africa before we even left our birth continent. Every passing month complicates the previous picture.

It’s likely to become more complicated still. Chris Stringer says that Currat and Excoffier’s study is an “important bit of work.” However, he thinks that their simulations are too simple. They assumed that modern humans spread from Africa to Eurasia in a single expanding wave. But Stringer says, “In my new book The Origin of Our Species I suggest that there were several waves of modern dispersal into Europe and no doubt several to the East as well.”

In the first waves, small groups of modern humans would have met large groups of Neanderthals. In later waves, the situation was reversed. The encounters between the two groups would have been very different across many thousands of years. Stringer asks, “The remaining question is whether those earliest modern waves survived to contribute their genes (including Neanderthal ones) to the succeeding waves, or was the slate essentially wiped clean each time?”

Reference: Currat & Excoffier. 2011. Strong reproductive isolation between humans and Neanderthals inferred from observed patterns of introgression. PNAS

Photo by Joe Dunckley

More on Neanderthal genes:


Comments (38)

  1. Great piece, Ed. This paper was one of mine as a now freelancer for PNAS. It’s nice to see the results so well explained in NERS. And how cool are these results?! Amazing that from the human genome we can figure information about the earliest humans’ sexual activities – or at least those that produced reproductive children.

  2. thanks for pointing to this!

  3. già da diverso tempo ho postato su Fb “gilgamesh la saga” – tra le mie note- alcune idee circa la fusione di cromosomi tra Neanderthal e i Cro-Magnon. La situazione e le scoperte sono in continua evoluzione .Sostengo in tali note che già 2-3000 anni a.c. l’uomo avesse la percezione della avvenuta fusione dei cromosomi, ovvero di rapporti sessuali tra le due specie. Infatti, nel racconto epico della storia di Gilgamesh ,le popolazioni mesopotamiche si tramandavano il racconto dell’avvenuto incontro tra le due specie :il cro-magno Gilgamesh (re di Uruk) e il neandertheliano Enkidu (uomo-bestia). Rileggiamo attentamente la Saga di Gilgamesh anche alla luce delle nuove scoperte scientifiche. Non dobbiamo dimenticare che nel racconto si narra altresì dell’incontro dell’uomo-bestia Enkidu con la prostituta con la quale ha rapporto sessuale. Ecco: eravamo già a conoscenza di tanti particolari da migliaia di anni ma non abbiamo capito!!

  4. Toos

    A really great overview of what’s going on in this matter [having my special interest by the way]. Thank you very much for this post!!!

  5. Isn’t it also possible that interbreeding was more extensive, but most Neandertal genes were ultimately selected against?

  6. Adriana

    Great article, as usual, Ed. I have a question (since I still can’t see the PNAS article for some reason, will keep looking for it). You say: “Even if the odds of successful interbreeding were just 5 percent, Neanderthal genes would make up the majority of the human genome today”. Why would that be, wouldn’t that depend on how large the human population was respect to the Neandertal population?

  7. @Jenny – Absolutely. I made this point in my New Scientist piece. Modern genomes can tell us a lot about the ancient stuff.

    @Mike – That would imply that the fraction that remains is all/largely adaptive, and there’s only a bit of evidence that some of it was adaptive.

    @Adriana – As to why the paper isn’t up yet, read this:

    As to your excellent question, the paper actually predicts an astonishing 80% introgression of Neanderthal genes at an interbreeding success rate of 5%. I asked Excoffier why. Here’s his reply (no type to parse this for general readers, but if anyone’s confused, just ask and I’m sure either I or someone else here will explain):

    This is a good question as this result is not intuitive, and is due to a combination of several factors. First, you need to realize that final introgression levels come from the accumulation of repeated interbreeding events occuring during the spatial expansion of modern humans into the Neanderthal territory. Moreover, these interbreeding events occur just on the front of the human expansion wave, where the two species still co-exist. However, these interbreeding events have a higher chance to occur when the two species have the same density. This time corresponds in our model at a point where the Neanderthal population is locally declining and the human population is increasing, which implies that if a Neanderthal gene enters the human population at this stage, it is is likely to be found in several copies when the human population has finished growing in size. With 5% interbreeding success, Neanderthal genes enter human populations not only in larger quantities but often at an earlier stage of the growing phase of the human populations on the front, which implies that they will also be more amplified than genes entering these populations later, leading to a non-linear increase of final introgression levels with interbreeding success.

  8. Why couldn’t most of the remaining bits simply not be maladaptive?

    (Just curious, I have no opinion of my own here.)

  9. Why “humans and neanderthals” as if neandertals weren’t humans?
    Why “humans” when you mean “sapiens”?
    Why the word “sapiens” is dissappearing from science writing?

  10. Yeah that whole thing is a minefield. You’ll note that I refer to Neanderthals, Denisovans and “other groups of ancient humans” in the first bit. It’s very hard to find the right term that gets across the fact that Neanderthals and modern humans were arguably the same species, while still being accessible to a lay reader. “Sapiens” doesn’t work either – if you decide that Neanderthals are the same species, then they’re Homo sapiens neanderthalis. In the end, I just decided to use “humans” to refer to modern humans, because that’s what the authors and many of the folks I interviewed did. And frankly, if it’s good enough for working scientists, I shouldn’t have to complicate it for a general reader!

  11. Thanks for your answer. It’s quite interesting that the scientists are avoiding the “s word”.
    In my opinion, if scientists and science writers keep on this new thing of using “humans” just for the _Homo sapiens_ species (or the _H. sapiens sapiens_ subspecies), this could cause a lot of confusion in the general reader.

  12. I’m not an expert on Neanderthal biology by any means, but I’ve been keeping up with the progression of the science — admittedly due to a bit of cynicism that Neanderthals and humans could interbreed at all.

    This particular study has me won over, but I’m not getting the idea that human-Neanderthal copulations were a “fringe” activity.

    Successful human pairings are already relatively rare. (Fertility clinics, anyone?) Successful interspecies (or inter-subspecies, perhaps) matings are going to be even more rare. And as stated in the article, lack of viable hybrids that can mate again might drop the success rate of DNA transmission even lower.

    Different gestational lengths would have made successful pregnancies even rarer. Humans have the 9-10 month gestation, while Neanderthals have a [theorized] 12-month gestation period.

    Putting all of the above together statistically is Above My Pay Grade. But I figure at least it means that the conclusion to draw from the resultant % of Neanderthal genetics present in humans is that interspecies sex happened fairly regularly within any individual’s lifespan.

    In other words, it doesn’t seem (to me) to be enough to say that the phenomenon of human-Neanderthal sex was simply *widespread*.

    “Considering that they two groups probably interacted for 10,000 years or so, it would have been enough for one human to sleep with one Neanderthal every 23 to 50 years.”

    So, humans and Neanderthals didn’t have much sex, AND these rare matings were often successful? Really?

  13. @El Paleofreak, restricting neontological terms (e.g., “human”) to crown groups is generally a good practice, as it limits unjustified inferences about stem groups. More discussion here:
    …and especially here:

  14. Alex

    Successful human pairings are rare? What? Wikipedia mentions that in sub-Saharan Africa, the average fertility rate is 8 children (, and anecdotally the fertility rate for certain religious groups like Orthodox Jews and Mormons is pretty high.

  15. it is hard enough getting pregnant as a human, nevermind as a neanderthals. imagine neanderthals were able to slog at

  16. Michelle

    Of course, you ignore that the wave may have gone in more than one direction. If it hadn’t humans in Africa would be far more different than humans elsewhere.

  17. “El Paleofreak, restricting neontological terms (e.g., “human”) to crown groups is generally a good practice”

    In this particular case I think it has too big disadvantages. And it’s too absurd: we have a genus, Homo, which literally means “human”. _Homo neanderthalensis_ means “Human from Neanderthal”.

  18. Actually, Naomi is right – we are a very low fertility species. As are all great apes (as far as we can tell, chimps, bonobos, orangs and gorillas are all very low fertility too). Many other animal species have sex once, or sex several times in a single cycle, and conceive… and not only conceive but have a whole litter of babies. We have on average a 40% chance of conceiving at any given cycle, and that’s assuming the timing of ovulation and endometrial receptivity are perfect. And then the majority of time we have a single baby.

    Add to that the fact that the vast majority of conceptions end in spontaneous abortion (very early pregnancy loss, before 6 weeks gestation), and you have lots and lots of cycles, and lots and lots of sex, for not a lot of babies.

    Given that a sub-Saharan forager woman will have about fifty menstrual cycles in her life (between when she is pregnant and nursing) and a post-industrial American will have as many as 400, there should be chances to have an awful lot of babies. But we don’t.

  19. Does the absence of Neanderthal mtDNA in humans not suggest that the surviving Neanderthal cDNA in humans is from male Neanderthals. That would also makes sense since the babies were probably more likely to stay with the mother.

  20. Torbjörn Larsson, OM

    Ooh, a _map_ of genome mixing.

    @ El PaleoFreak:

    But now you insist of imputing meaning where there currently is none.

    There is a lot of alike (homo) groups out there. The question is how alike (us) they are.

    I am sure the terminology will be straightened out to a comfortable level as soon as the species/subspecies and human qualification issues are less fluid.

    @ KateClancy:

    Actually, Naomi is right – we are a very low fertility species. As are all great apes (as far as we can tell, chimps, bonobos, orangs and gorillas are all very low fertility too). Many other animal species have sex once, or sex several times in a single cycle, and conceive… and not only conceive but have a whole litter of babies.

    I think you are comparing apples, and oranges, with pears here.

    – Number of offspring is an ecological strategy. Individual lifetime too. (Apples of great apes.)

    – As for humans we have cryptic ovulation (“cycles”), which works fine for a societal species to keep aggression down and mates around if it _isn’t_ highly fertile, else not. (Oranges of humans.)

    – Conversely if you ovulate once a year for a few years, it better take. (Pears of most mammals.)

    Besides, “sleep” were referring to successful sex here. I think the article and comments explains why low attempt frequency rather than low success rate are the suspect here. (From population genetics considerations.)

  21. Eleanor

    @Bjorn (sorry, can’t find the accent on my keyboard)

    They model the predicted intogression of Neandertal mtDNA under the best supported scenarios, and never recover any at the end (modern populations). The numbers of Neandertal mtDNA haplotypes are so small they just get lost by the time we get to the present. So no data in support of male humans/ female neandertals or vice versa!

  22. Torbjörn Larsson, OM

    @ Michelle:

    Of course, you ignore that the wave may have gone in more than one direction. If it hadn’t humans in Africa would be far more different than humans elsewhere.

    From my armchair I don’t see how that follows.

    First, the original colonization wave happened from advancing spreading centers, even though there can be canalization (to coasts and islands, say). But a wave front of any geometry can be composed of exactly such spreading centers by Hyugen’s principle.

    Whether you count that as “one direction” or a normal spreading wave I don’t know. Waves are collective phenomena with little localization, so “direction” and separation (say, as breeding) doesn’t mean much for a local volume. You would have to dig into the actual dynamics to tease out any dissimilarities to the more basic wave phenomena.

    Second, after a while the initial wave into land uncolonized by a certain subspecies (modern humans) would be difficult to separate out of new migratory waves.

    Third, the already colonizing subspecies (neanderthals et cetera) would be migrating and mixing internally and with each other too. Do we know how much of this was going on?

    Fourth, mixing would happen also in Africa. There mixing is currently suspected to have happen (from patterns of “ancient DNA” AFAIU), presumably with remaining neanderthals and what have you.

    Fifth, a mere 0.2 Ma for modern human as a subspecies as opposed to the 0.5 – 1 Ma for subspecies splits would mean modern humans would be more homogeneous than the genomes material it mixed with during “the wave”. That homogeneous material composed the dominant part.

    And it would be little time to establish absolute differences, some 0.1 Ma or less since the “Escape Out of Africa”. Didn’t lactose tolerance take some 5-10 ka to appear and spread through such populations? And that was a huge selective advantage.

    Sixth, isn’t checking the predicted level of mixing what the paper asks for? “Our model additionally suggests that similarly low levels of introgression in Europe and Asia may result from distinct admixture events having occurred beyond the Middle East, after the split of Europeans and Asians. This hypothesis could be tested because it predicts that different components of Neanderthal ancestry should be present in Europeans and in Asians.”

    So there may or may not be relative “far” differences, for a number of reasons. But it doesn’t seem to be ignored but crucial.

  23. @Torbjorn, I am going to be honest here and say I have absolutely no idea what your response to my comment means.

  24. It just seems like the beautiful vistas of the Mediterranean got them all frisky.

    The Med and… Norfolk!

  25. Adriana

    Thanks, Ed, for posting what the authors replied to your question. Wow; an 80% introgression rate? I understand their explanation, but of course these models are all based on assumptions, especially that interbreeding must have occurred very early on.

  26. Heleen

    The John Hawks link doen’t work for me.

  27. chris y

    Currat and Excoffier suggest that either modern humans and Neanderthals didn’t have sex very often, or their hybrids weren’t very fit. They favour the first idea.

    Speculation about the behaviour of people who lived two thousand generations ago aside, do they explain why? I’d have supposed that it would be quite likely that interbreeding between two populations that had been separated for so long would produce very few viable offspring, or at least fertile ones. This is what we tend to see in other genera, surely.

  28. Mia

    I have a question for you, Ed (well done, by the way!):

    Is it necessarily true that they had few sexual encounters or few surviving offspring?

    Is it possible that the sexual encounters were common -and- there were many offspring, BUT a majority of the hybrid offspring were sterile and thus, very few Neanderthal-human offspring went on to reproduce?

    Would that alternative theory work?

    Really interesting article!

  29. Karen

    You mention Denisovans and other ancient ‘humans’ at the beginning of the article. Could we not have gained our Neanderthal DNA via third party – i.e., a descendent from a Neanderthal/Denisovan pairing subsequently mating with a homo sapien?

    I find the topic fascinating and your article enlightening from a geographic perspective. You may want to find a new Proofreader who’ll catch those typos, though.

  30. BalRog

    About the lack of genetic evidence for successful human-male-neanderthal-female paring, two ideas occur to me. First is the species-ist idea that human women were more attractive to both sets of men. This idea clearly arises from the same base part of the mind that has aliens kidnapping human women on innumerable old sci-fi pulp magazines.

    The second idea is more of an image really. I am imagining a baby homo sapiens cranium trying to fit though a homo neanderthalensis birth canal. I’m not imagining a happy result for either mother or child.

  31. It would have surprised me much more if humans and Neanderthals did not have sex. A quick scan of the Internet will show you humans having sex with dogs, horses, cows, pigs, octopodes, goats, sheep, bears, gorillas, liver, dead people, zucchinis, carrots, cucumbers, beer bottles, watches…

  32. Il Cro-magnon non esiste più ? Chi dovrebbe essere ? In questo commenti non viene mai citato !

  33. Bob Dole


    “The second idea is more of an image really. I am imagining a baby homo sapiens cranium trying to fit though a homo neanderthalensis birth canal. I’m not imagining a happy result for either mother or child.”

    Um… Neanderthal skulls were larger than human skulls, so this doesn’t make too much sense.

    Source: “The female Neanderthal pelvis was wider than in modern human females–but the head size of Neanderthal newborns was also larger.”

  34. MartinP

    Isn’t it possible or even likely that more mixed-blood children were born, and that they were perfectly capable of reproducing, but not allowed to because they were considered ugly, shameful, etc? And it follows that the lack of modern human genes in Neanderthals could be explained by the discrimination being stronger in neanderthal society, to the point of even leaving a baby out to be eaten by the jackals.

  35. Richard W. Crews

    Didn’t the British Navy, during the 1800’s, prove that a man will f*ck anything/anyone? It’s human nature!
    And, could it be possible that the hybrids were victims of racial bigotry, and worse than not being able to get a date, were shunned, when that could well have been a death sentence? A non-progenic (sp?) life with no heirs?
    Back in the 50’s TV Disneyland would talk about the Future, showing egg-headed bald guys with big button-pushing fingers. Even at age 12 I knew this was incorrect; those guys wouldn’t get married and pass on anything. I was partly wrong because of the power of money – ugly rich guys can rule.

  36. Enjoyed the article and the many responses. I am involved in some archaelogical work here in the west of Turkey which I notice is a bit of a ‘hot spot’
    I have seen the way the young waiters chase anything in a skirt no matter age/looks/ethnic origin if that helps the discussion!

    To Richard Crews, may I point out that with my long departed father a Royal Navy officer with a distinguished career of over forty years and his line of serving ancestors that go all the way back to around 1780 also in the Navy, I can assure him that unless it was a lush exotic South Seas fruit on a lonely night, it otherwise had to at least to be breathing!

  37. Felton

    So could the Neanderthal in Europeans explain the light skin. I understand the theory of climate being the rationale for this variance, but maybe it was simply bred into Europeans.

  38. Corin Price

    I’m sorry to be a bother, but this is starting to piss me off – we share 98.4 percent of our DNA with our closest living relatives, no one seems to know how much DNA cro-magnon & neanderthal shared before the mingling of genetics, how closely related were the 2 early species I cannot find a reference anywhere – I mean seriously, how closely related was their DNA before, I don’t want to know about now, I want to know about then (if I could figure out how to italicize I would)


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar