Few origin claims in Europe are argued as fiercely, or with as much political charge, as the Albanian one. Albanians describe themselves as the last direct heirs of the Illyrians, the Iron Age peoples of the western Balkans who resisted Rome and then, unlike their neighbours, were never wholly absorbed by the Slavic migrations that reshaped the peninsula after the sixth century. Critics have countered with two rival stories: that Albanians are not specifically Illyrian at all but a broader paleo-Balkan population from the inland Dardanian and Daco-Thracian contact zone, or, in the strongest revisionist version, that they are relative newcomers, a medieval wave that entered the region only around the eleventh century. For most of the last two hundred years these three positions could only be argued from language, place-names and scattered chronicles. They can now be tested directly, because the dataset used here includes something rare and decisive: an in-situ genetic transect from a single northern Albanian valley that runs from the Bronze Age through the Iron Age Illyrian horizon and on into the medieval, early modern and living population.
Three Hypotheses, One Testable Valley
The three claims make different, checkable predictions. The Illyrian-continuity thesis predicts that modern Albanians should sit close to the Iron Age populations of the western Balkans and, ideally, show an unbroken local line of descent in a place they have demonstrably always lived. The broader paleo-Balkan thesis accepts the deep-rootedness but denies that any single tribe, and the coastal Illyrians in particular, can be crowned the ancestor; it predicts that Albanians should sit more or less equally close to several Iron Age Balkan groups rather than to one. The medieval-wave thesis predicts the opposite of both: a genetic profile that only assembles late, with a large and recent eastern or Slavic component and no meaningful line back to the local Iron Age at all.
The material that lets us adjudicate them is the Cinamak series from the Kukes region of north-eastern Albania, a run of dated genomes from essentially the same place spanning the Late Bronze Age, the Late Iron Age Illyrian period, the Roman Republican transition and beyond, supplemented by early medieval, late medieval and early modern Albanian genomes and by paleo-Balkan comparanda from Montenegro, Serbia and Croatia. Against that backdrop, four questions can be answered in turn: where modern Albanians sit relative to the Iron Age Balkans and the Slavs, whether the local line is genuinely continuous, how much of the Slavic wave actually arrived, and whether the surviving signal is specifically Illyrian or broadly paleo-Balkan.
Modern Albanians Sit With the Iron Age Balkans, Not the Slavs
The first and bluntest test is a distance ranking. Taking modern Albanian (Tosk) as the reference and ranking every relevant ancient and modern population by G25 Euclidean distance from it, the entire top of the list is paleo-Balkan. Early medieval Albania, Iron Age Serbia, Late Bronze Age Montenegro and the Iron Age Illyrian genomes of the Cinamak valley itself all sit within a tight band between 0.029 and 0.036. The Slavic pole, represented here by an early medieval population from Poland, sits far away at 0.132, roughly four times more distant than the Iron Age Balkans. Whatever modern Albanians are, they are not a lightly-repackaged Slavic population.
This is exactly the shape the medieval-wave hypothesis predicts against. A population that had genuinely arrived from outside in the eleventh century, or that was mostly Slavic in origin, would sit closest to the Slavic pole and far from the local Iron Age. The observed pattern is the reverse: the Iron Age Balkans are the near neighbours, and the Slavic reference is the outlier. The Bronze Age of the same region sits a little further out at 0.055, which is expected and even reassuring, since two and a half thousand years of gradual gene flow separate the Bronze Age from the present, and a modern population should sit closer to the Iron Age than to the deeper Bronze Age of its own homeland.
A 2,500-Year Line in One Albanian Valley
Distance rankings show affinity, but they do not by themselves prove continuity, because affinity to a regional average can be assembled from several directions. The Cinamak transect closes that gap, because it lets us watch a single locality across time and ask whether it stays put. Anchoring on the Late Iron Age Illyrian genomes from Cinamak and measuring the distance to every other stage of the same valley produces a strikingly flat line.
Every phase of the valley, from the Middle to Late Bronze Age transition, through the Late Iron Age Illyrian horizon, into the Roman Republican period, the medieval Kenete and Barc genomes, the early modern Bardhoc population and finally the living Albanian average, stays within roughly 0.03 to 0.05 of the Iron Age anchor. There is drift, but it is small and gradual, of exactly the magnitude two and a half millennia of ordinary local demography would produce. There is no break, no replacement horizon, no sudden lurch toward the steppe or the Slavic pole at any point in the sequence. This is about as literal a demonstration of population continuity as archaeogenetics can offer: the people burying their dead in this corner of north-eastern Albania in the Iron Age are, to a very good approximation, the direct ancestors of the people living there now.
How Much of the Slavic Wave Actually Arrived?
Continuity does not mean isolation. The Slavic migrations were real, and Albanian, like every Balkan language, carries Slavic loanwords; the honest question is not whether a Slavic layer exists but how large it is. A two-source NNLS model, built from a local medieval Albanian anchor and a clean Slavic pole, quantifies it directly, and the results are both modest and neatly structured by geography.
The southern and core Albanian groups take on the least: the Cham of Greece and the general Tosk population sit at 8 to 12 percent Slavic, with Gheg close behind. The share rises as one moves north and east into the zones of heaviest medieval Slavic settlement, reaching 17 percent in North Macedonia, 18 percent in Kosovo and 23 percent among the Gheg of Montenegro. Even the highest Albanian figure, though, is comfortably below that of the surrounding Slavic-speaking populations modelled the same way: Macedonians come out at 26 percent, Bulgarians at 34 percent and Serbs at 47 percent on the same scale. The gradient is coherent and it points the right way. Albanians absorbed a real but limited Slavic layer while remaining, throughout, a majority paleo-Balkan population, and they did so least in precisely the regions their tradition treats as the old heartland. These figures line up closely with the published estimates in the 2026 Nature Human Behaviour study, which put the later medieval eastern European admixture into Albanians in the same modest range.
Specifically Illyrian, or Broadly Paleo-Balkan?
Here the data forces a genuinely important nuance, and it is one the more triumphalist versions of the Illyrian claim tend to skip. Modern Albanians are unambiguously paleo-Balkan, but the autosomal signal is not narrowly Illyrian. When the same distance measure is turned on a panel of Iron Age Balkan anchors, the modern population sits about equally close to all of them.
Inland Iron Age Serbia at Gomolava (0.030), Late Bronze Age Adriatic Montenegro (0.032), the coastal-facing Illyrian genomes of Cinamak (0.034 to 0.036) and north-western paleo-Balkan Croatia (0.036) are all bunched within a hair of one another. On the autosomal evidence alone there is no basis for crowning the coastal Illyrians specifically, rather than the inland Dardanian and Daco-Thracian populations, as the single ancestor. This matters because it reconciles the genetics with a long-standing linguistic objection: Albanian borrows its maritime vocabulary from neighbouring languages rather than inheriting it, which has always suggested that the language did not form on the coast but somewhere inland, in exactly the Dardanian contact zone where Illyrian and Daco-Thracian speech met. The genome cannot resolve coastal Illyrian from inland Dardanian, because in G25 space they are barely distinguishable to begin with; what it can say is that the ancestry is paleo-Balkan through and through.
The specifically Illyrian pin, where it exists, comes not from the autosomes but from the male line. The paternal lineage most associated with modern Albanians, J2b-L283, is found in Bronze and Iron Age western Balkan contexts and has a remarkably local pre-Roman distribution, which is why the 2023 ancient-DNA study of Albanian origins could describe a genuine paternal continuity from Bronze Age Balkan populations including those known as Illyrians. The fair summary is therefore a layered one: broadly paleo-Balkan by the genome as a whole, with a specifically west-Balkan and plausibly Illyrian signal carried most clearly by the Y chromosome.
The Medieval-Wave Hypothesis Does Not Survive
The oldest and most politically loaded of the three claims, that Albanians are a late medieval arrival with no deep local roots, is the one the data dispatches most cleanly. It was never a strong hypothesis to begin with, resting largely on the late first appearance of the Albanian name in written sources in the eleventh century, but the genetics leave it no room at all. A population assembled by a medieval migration would show its profile snapping into place only in the medieval period, with the pre-medieval locals as strangers. Instead the Cinamak transect shows the modern profile already essentially present in the Iron Age and holding steady across every intervening century, and the recent peer-reviewed work on Balkan population history reaches the same conclusion independently, finding the Albanian genetic profile already established by around 800 to 900 CE with only limited later eastern European admixture on top. The written record is simply silent for longer than the genome is; absence of a name is not absence of a people.
Limits and Caveats
Several caveats apply and are worth stating plainly. The Cinamak transect, powerful as it is, rests on a modest number of individuals per phase, and a single valley cannot stand in for the whole of ancient Albania; it is a very good local record, not a national census. The paleo-Balkan anchors used in the fourth test are close enough together in G25 space that small differences in their distances should not be over-read, which is precisely the point being made with them rather than a flaw to be explained away. The two-source Slavic model is a directional estimate: it deliberately reduces a complex history to a single local-versus-Slavic axis, and the exact percentages should be treated as approximate, with the plain distance rankings weighted more heavily than any single NNLS figure, as throughout this site. The autosomal method genuinely cannot separate coastal Illyrian from inland Dardanian ancestry, so the article does not claim it can; the Illyrian-specific reading depends on the paternal-lineage evidence from the published literature, not on the coordinates modelled here. Finally, the deeper linguistic question of where exactly Albanian formed is not a genetic question at all, and the genome is consistent with, but cannot by itself prove, the inland origin the loanword evidence points to.
Conclusion
Of the three hypotheses the title poses, one is dead and two are really one. The medieval-wave story does not survive contact with an Iron Age to modern transect that never breaks. The Illyrian-continuity and broad paleo-Balkan stories turn out to be the same story viewed at different resolutions: modern Albanians are a deeply rooted paleo-Balkan population, continuous in place for at least two and a half thousand years, carrying a real but modest and geographically sensible Slavic layer of roughly 8 to 23 percent, and bearing their most specifically Illyrian signal on the male line rather than across the whole genome. They are not a coastal Illyrian population to the exclusion of their inland neighbours, and they are certainly not medieval newcomers. They are, on this evidence, about as convincing a case of long-term paleo-Balkan continuity as the peninsula has to offer, which is very likely why the claim has been worth fighting over for so long.
Albanian_Tosk_(n=42),0.11902645,0.14405882,0.01415819,-0.020190795,0.022837452,-0.0077694952,0.0022493095,0.00067580952,0.001972119,0.017327643,0.00080699048,0.0035503571,-0.0058585524,0.0047185,-0.018907119,-0.0028890476,0.010623174,0.0008325,0.0071588333,-0.0056157857,-0.0096258333,0.0014956429,0.0026470071,0.0029779762,-0.0040174348
Albanian_Tosk_Cham_Greece_(n=4),0.118376,0.1457285,0.01376475,-0.0237405,0.021773,-0.00732075,-0.00088125,-0.001096,0.001943,0.02086625,-0.0003245,0.00651925,-0.0116325,0.00199525,-0.01394525,0.000033,0.014114,0.00307225,0.009019,-0.002345,-0.00932725,0.00034,0.0013865,0.005302,-0.00179625
Albanian_Gheg_(n=46),0.12502433,0.14743948,0.0163265,-0.021333326,0.026001,-0.0093016957,0.0023833261,0.00069528261,0.00069930435,0.020265152,-0.00021604348,0.0041951957,-0.0071743696,0.0081692174,-0.019371022,-0.0045188261,0.0099716087,0.0012826522,0.00833695,-0.0080202391,-0.0097691957,0.0012951304,0.0019258261,0.0045185652,-0.0037701304
Albanian_(n=13),0.12150923,0.14373646,0.014823769,-0.018388538,0.025732538,-0.0076586923,0.0043565385,0.0036743846,-0.00029884615,0.018558154,0.0012740769,0.0022134615,-0.0067583846,0.0068283077,-0.020326923,-0.0067008462,0.0086153846,0.00074061538,0.0062172308,-0.0063876154,-0.0067283846,0.00040884615,0.0021615385,0.0022523077,-0.0034910769
Albanian_Gheg_North_Macedonia_(n=11),0.12355355,0.14329625,0.021500455,-0.017162836,0.026091727,-0.0061642727,0.0023329091,0.00070536364,0.0021175273,0.018379991,-0.00043663636,0.0037871636,-0.0058400364,0.007217,-0.015580055,-0.0040883636,0.011428973,0.0013835455,0.00684,-0.0074740273,-0.0075737273,0.0011159455,0.005523,0.0019802182,-0.0021761455
Albanian_Gheg_Kosovo_(n=54),0.12479239,0.14563498,0.019777667,-0.013549463,0.026465537,-0.005356537,0.0038550926,0.0012128148,0.0015589444,0.017337019,0.00011948148,0.0035290741,-0.005771463,0.0081608889,-0.017946685,-0.0041111481,0.0095997778,0.00031774074,0.0087359815,-0.0068590741,-0.010469778,0.0013533148,0.0029516852,0.0028974259,-0.0022041111
Albanian_Gheg_Montenegro_(n=9),0.12586989,0.14262867,0.023194444,-0.0099583333,0.026968889,-0.0055181111,0.0029768889,0.0019346667,0.0038993333,0.014905667,-0.0022078889,0.00015333333,-0.0034933333,0.011502667,-0.017396111,-0.0068982222,0.0039183333,0.0016472222,0.0061608889,-0.0092204444,-0.010412889,-0.00043377778,0.003684,-0.00021066667,-0.0024491111
Macedonian_North_Macedonia_(n=47),0.12442898,0.1421386,0.024031106,-0.0066676383,0.025994468,-0.0026295957,0.003327234,0.0021624468,0.00046470213,0.0097982766,0.000805,0.0016527872,-0.0017363617,0.0095140638,-0.014937936,-0.0037887234,0.0050865957,0.00067844681,0.0052293191,-0.0045183191,-0.0076537234,0.0003606383,0.0032962128,-0.000016382979,-0.0012450638
Bulgarian_(n=16),0.12321344,0.13500175,0.025879875,0.001615,0.023138875,0.001568625,0.0056253125,0.0044996875,-0.0025309375,0.005398625,-0.0004465625,-0.000946,0.0010221875,0.010984125,-0.013495687,-0.0040521875,0.0060301875,0.000807625,0.00323675,-0.0028294375,-0.0059895,0.000865625,0.003019625,0.000715375,-0.00199825
Serbian,0.127311,0.137587,0.039943,0.015698,0.032068,0.004895,0.004653,0.006411,0.000508,0.001488,-0.001378,-0.001474,0.002859,0.012297,-0.013746,-0.001772,0.00583,0.000289,0.00591,-0.002357,-0.007726,-0.000998,0.00502,-0.000207,-0.001579
Albania_MBA_(n=1),0.124067,0.144205,0.017725,0.000323,0.03416,-0.019801,0.00564,-0.012461,0.011658,0.034078,0.003248,0.004346,-0.005352,-0.022432,-0.008822,0.004243,0.026989,-0.007475,0.001885,0.006128,-0.000873,0.006059,0.007765,0.010724,0.008143
Albania_MLBA-LIA_Cinamak_(n=1),0.125205,0.149283,0.034318,-0.00646,0.029544,-0.005857,-0.001175,0.004846,-0.00225,0.019681,0.001461,0.002698,-0.020069,-0.006468,-0.003122,0.002121,0.001825,0.003167,0.002011,-0.002501,-0.007861,0.003462,0.005423,-0.004217,-0.005269
Albania_LBA-LIA_Cinamak_(n=1),0.125205,0.152329,0.025644,-0.010336,0.019388,0.001116,-0.008695,0.003461,0.006136,0.025878,0.007145,0.010641,-0.018583,-0.00234,-0.011536,-0.006099,0.008605,0.00038,0.009679,-0.004877,-0.010981,-0.006183,-0.001479,0.012411,0.003113
Albania_LIA_Cinamak_(n=4),0.1232135,0.1566455,0.030924,-0.01267775,0.0323135,-0.0073905,0.003995,-0.00340375,-0.00153375,0.02523975,0.00263875,0.006894,-0.0180995,-0.0067435,-0.009093,-0.00245275,0.0149615,0.002407,0.00282825,-0.01006725,-0.005116,0.0024115,-0.005084,0.009911,-0.00793325
Albania_LIA-Roman_Republic_Cinamak_(n=1),0.124067,0.144205,0.018102,-0.019703,0.025851,-0.003347,0.002115,0.008077,0.000205,0.022233,-0.000974,0.005095,-0.017096,-0.017753,-0.013301,-0.003447,0.00665,-0.003294,0.010182,-0.013006,-0.000125,0.005441,-0.007148,-0.001687,-0.004431
Albania_Early_Medieval_Kenete_(n=1),0.111547,0.15436,0.013953,-0.033592,0.016003,-0.001673,-0.001645,-0.001385,-0.001227,0.016037,0.006008,0.005845,-0.005946,0.00234,-0.013436,-0.013259,0.015255,-0.007728,0.009302,-0.012381,-0.015348,0.006801,-0.005669,-0.00735,-0.005389
Albania_Early_Medieval_Shtike_(n=1),0.120652,0.151314,0.013953,-0.037468,0.030159,-0.012271,-0.001175,-0.001615,0.006136,0.025149,-0.004222,0.003897,-0.015609,0.002202,-0.019815,-0.007425,0.008605,0.006461,0.00817,-0.014257,-0.007112,0.002968,0.006039,-0.003735,-0.000599
Albania_Late_Medieval_Barc_(n=1),0.12862,0.145221,0.01961,-0.01292,0.012925,-0.007809,-0.00141,-0.006461,0.002863,0.025878,-0.000325,0.006145,-0.022002,-0.004266,-0.012486,-0.009546,0.003912,0.003421,0.008045,-0.013131,-0.010107,0.006059,-0.004437,-0.000602,-0.004191
Albania_Early_Modern_Bardhoc_(n=4),0.12207525,0.1518215,0.01329375,-0.02365975,0.027928,-0.00836675,0.0017625,0.00207675,0.0017385,0.02132175,0.00215175,0.0049085,-0.01148425,0.00350925,-0.022428,0.00152475,0.02219775,0.00148875,0.007102,-0.00603425,-0.0141625,0.00550275,0.002095,0.003615,-0.00637675
Montenegro_LBA_(n=11),0.12706782,0.14919036,0.027872636,-0.0069885455,0.027333727,-0.00573,-0.00057681818,0.0010698182,0.0039602727,0.029207727,0.0040745455,0.0097004545,-0.01342,-0.006781,-0.010462818,0.000036,0.006377,0.00046072727,0.0056793636,-0.0089245455,-0.0069195455,0.0050473636,-0.00048181818,0.0054113636,-0.0012844545
Serbia_EIA_Kalakaca_Gomolava_(n=17),0.12928971,0.15262812,0.027840412,-0.011134,0.033327412,-0.0077104706,0.0012164706,0.00081447059,0.0081932353,0.024644765,0.0014711176,0.0060827647,-0.014288882,-0.00028335294,-0.010322706,-0.0037827647,0.0055223529,0.0011102353,0.0077192941,-0.0032516471,-0.0070831176,0.0051061176,0.00073947059,0.0056208235,-0.0016835882
Croatia_Late_Antiquity_Bubi's_Cave_(Paleo-Balkan_Profile)_(n=18),0.12880978,0.14623617,0.032411389,-0.0059396111,0.032279444,-0.0052679444,0.0020498333,0.0019616111,0.0071582778,0.025816778,-0.0015608333,0.0053950556,-0.015650556,-0.0042356667,-0.0063411667,-0.0063127222,0.0052152222,-0.00011255556,0.0026187222,-0.0072603889,-0.0026827778,0.0037507222,0.00017116667,0.0020886111,-0.0038186111
Greece_Classical_Arta_Ambracia_(n=5),0.1224736,0.159032,0.0015084,-0.0541348,0.0249892,-0.0203034,0.001739,-0.0061382,0.0066674,0.0399824,0.001981,0.0095316,-0.0133794,-0.0008806,-0.0229096,-0.002148,0.014655,0.0002026,0.0073408,-0.0097048,-0.0133264,-0.000099,0.0025882,0.0073262,-0.0020594
Poland_Early_Medieval_Niemcza_(Balto-Slavic_Profile)_(n=8),0.1253475,0.12389475,0.075094375,0.067951125,0.043661875,0.031479875,0.01110425,0.012374375,-0.006365875,-0.025148375,-0.002253,-0.00683775,0.011093875,0.02442825,-0.014047,-0.00285075,0.002624125,0.002011125,-0.0000315,0.006081,-0.0008735,0.0020555,0.006008375,-0.007937875,0.00374225
- Davranoglou et al. Ancient DNA reveals the origins of the Albanians, bioRxiv preprint, 2023. biorxiv.org/content/10.1101/2023.06.05.543790v1
- Nature Human Behaviour Ancient DNA evidence for the history of the Albanians, peer-reviewed, 2026, on a paleo-Balkan profile established by around 800 to 900 CE with limited later medieval eastern European admixture. nature.com/articles/s41562-026-02462-z
- Olalde et al. A genetic history of the Balkans from Roman frontier to Slavic migrations, Cell, 2023. cell.com, Olalde et al. 2023
- Davidski Global25 coordinates dataset.
- Vahaduo G25 analysis tool used for NNLS modelling.
- Moriopoulos 2026 collection Aggregated Global25 population averages, including the Cinamak Bronze Age to Roman transect, the medieval and early modern Albanian genomes, and the Montenegro, Serbia and Croatia paleo-Balkan comparanda used in this article.