The Genetic History of the Romani People

The Romani people are the most studied diaspora population in European population genetics, and the result of that scrutiny is one of the most precisely dated migration events in the modern human record. Genome-wide analyses converge on a single conclusion: the Romani originate from a small founding population that left northwest India approximately 1,500 years ago, passed through a severe demographic bottleneck, traversed Persia and the Byzantine Empire over several centuries, and entered the Balkans by the late thirteenth or early fourteenth century. The descendants of this founding population spread across Europe and have absorbed substantial European admixture in each region, but the South Asian genetic and linguistic signal has remained recoverable everywhere. The Romani language, classified as Indo-Aryan with closest affinity to the Central Indian zone, is the linguistic record of the same migration. Y-chromosome haplogroup H1a1a-M82, mitochondrial haplogroups M5a1 and M18, and a battery of identity-by-descent shared segments all point back to the same founding event. Modern Romani populations of Europe are, on average, 20 to 30 percent South Asian and 70 to 80 percent European in autosomal terms, with strong regional variation reflecting where they settled and how endogamous they remained. This article reconstructs the genetic, linguistic, and historical record of a population that has become genetically European on the surface while preserving, in striking and recoverable detail, the founder signature of a medieval Indian emigration.

1. The single migration: what genome-wide data revealed

The Romani case is unusual in population genetics because almost everything about it was contested for centuries on the basis of partial linguistic, historical, and ethnographic evidence, and almost all of it has now been settled by genome-wide data. The basic question, whether the Romani are a single diaspora descended from one founding migration or a heterogeneous collection of populations that converged on a shared identity, was answered definitively by Mendizabal and colleagues in 2012. Using dense SNP genotyping across thirteen Romani populations from six European countries, they showed that all sampled European Romani populations descend from a single founding event, with no evidence of multiple independent Indian origins. The founder signal was visible in shared IBD segments, in a common signature of bottlenecked allele frequencies, and in the consistent Y-chromosome and mitochondrial haplogroup profile across geographically distant Romani communities.

This finding put to rest a long line of alternative hypotheses. The Romani are not descended from Egyptians (despite the etymological origin of the English word Gypsy from Egyptian, a medieval European misidentification). They are not descended from a heterogeneous mix of Middle Eastern and South Asian peoples gradually accreted into a shared identity. They are not the result of multiple independent migrations from different parts of India at different times. The genome says one migration, one founding population, one Indian source region, and a tight founder bottleneck.

2. The Indian origin: northwest India, specifically

The South Asian component of Romani ancestry has been narrowed by successive studies to a specific region of northwest India. Moorjani and colleagues (2013, PLoS Genetics) used haplotype-sharing analysis to test which modern Indian populations most closely match the Romani founder signature. The strongest matches came from populations of the Punjab, Rajasthan, Haryana, and Gujarat, in the northwestern quadrant of the Indian subcontinent. The match was less close with central and southern Indian populations, ruling out a Dravidian South Indian origin.

The Indian endogamous structure further narrows the picture. Within the northwest, populations grouped under historical caste categorizations such as Doma (musician and laborer communities), Lohar (blacksmiths), and Banjara (itinerant traders) show some of the closest haplotype matches to Romani founders. This is consistent with the linguistic and historical interpretation of the Romani as descendants of an itinerant or service-caste population from northwest India that crossed into Persia in the second half of the first millennium.

The G25 coordinates below place the Romani diaspora communities of Europe alongside the Northwest Indian source populations identified by Mendizabal et al. 2012 and Moorjani et al. 2013 as the most likely region of origin. Modern Roma populations cluster between their European host populations and the South Asian source signal.

Roma_Balkans_Bosnia,0.091059,0.063640,-0.058831,0.024979,-0.018772,0.013108,0.000392,0.002539,0.016703,0.014882,0.000379,-0.000350,0.000942,0.008762,-0.010541,-0.005215,0.008866,0.003125,0.001508,0.001001,-0.002080,0.000124,-0.000657,-0.004016,-0.001277
Roma_Czechia,0.087419,0.059210,-0.056244,0.024590,-0.017033,0.013238,0.001951,0.004357,0.011212,0.012636,-0.000824,0.000102,-0.000492,0.003162,-0.006689,-0.001236,0.002694,0.001042,0.001030,-0.002616,-0.002766,0.000621,0.001211,0.002081,-0.000701
Roma_Serbia,0.101303,0.088351,-0.014331,0.017765,0.007386,0.006693,0.001175,0.004154,0.003068,0.012210,-0.003735,0.002248,-0.001041,0.005780,-0.012758,0.009016,0.018515,0.004181,0.005908,-0.002001,0.004367,-0.000371,0.004560,-0.000482,-0.005748
Roma_Spain_Barcelona,0.101113,0.099691,-0.007417,0.014158,0.011336,0.009296,0.000235,0.003115,0.018476,0.026121,-0.000704,0.002897,-0.008399,-0.002454,-0.002217,-0.004044,0.003847,-0.002154,-0.001027,0.000021,-0.000416,-0.001443,-0.003410,-0.006005,-0.000040
Roma_Spain_Bilbao,0.090490,0.068929,-0.045773,0.020268,-0.003000,0.017187,-0.000118,0.002308,0.019277,0.021595,-0.001401,-0.000019,0.000167,0.006967,-0.008500,-0.009331,-0.000896,0.000649,-0.000189,0.000516,-0.002090,0.001144,0.001017,-0.002094,0.000569
Roma_Spain_Granada,0.082278,0.068040,-0.053551,0.018088,-0.010771,0.014343,0.001007,0.006066,0.021504,0.019421,-0.000394,0.000792,-0.005394,0.004561,-0.008647,-0.005398,0.002049,-0.003837,-0.002011,0.000143,-0.001604,-0.001696,-0.000933,-0.002995,0.004858
Roma_Spain_Madrid,0.089067,0.072610,-0.049780,0.020026,-0.006771,0.010389,0.005699,0.002077,0.014317,0.021778,0.003613,-0.004159,-0.004125,0.002306,-0.002171,-0.002983,0.003455,0.000570,0.000691,0.000532,0.000905,0.003926,0.004437,-0.001085,0.003113
Roma_Portugal_Porto,0.086506,0.071595,-0.050534,0.021883,-0.007463,0.018058,0.002585,0.003750,0.016311,0.021732,-0.000893,0.002323,-0.005426,0.005092,-0.010451,-0.008585,0.000163,-0.002376,0.000315,0.006628,-0.002122,-0.003586,0.001941,0.002591,-0.003114
Roma_Turkey_Balkans,0.091058,0.057885,-0.078818,0.019380,-0.022773,0.015897,-0.003290,0.003923,0.020248,0.020410,-0.000487,-0.001349,0.000892,0.012937,-0.013843,-0.001856,0.011735,0.001774,0.003142,0.007379,-0.002121,-0.003462,-0.001972,-0.003374,-0.000838
Punjabi_Sikh_India,0.07275,0.000757,-0.10378,0.080723,-0.070121,0.047708,0.001635,0.004399,-0.000844,-0.005417,-0.005763,-0.000172,-0.000348,-0.00708,0.010373,0.008489,-0.003945,0.001084,0.000773,-0.010388,-0.002296,-0.00547,0.000857,-0.000718,0.000428
Brahmin_Punjab,0.068294,-0.00914,-0.108611,0.084949,-0.070475,0.055778,0.000705,0.005769,0.000409,-0.005103,-0.006333,0.003747,-0.00773,-0.011423,0.005429,0.015778,0.007041,0.002027,0.005154,-0.00963,0.000873,-0.001978,0.00419,0.001807,-0.000359
Rajput_Rajasthan,0.060779,-0.020311,-0.129741,0.091751,-0.066011,0.053206,-0.002939,0.010384,0.016768,0.010023,-0.007226,-0.000449,-0.001189,-0.000523,0.001724,0.002453,-0.007302,-0.00158,-0.001634,-0.002001,-0.001873,-0.003709,-0.001725,-0.001988,-0.002035
Jat_Haryana,0.066303,-0.017756,-0.121483,0.088126,-0.068013,0.056196,0.001657,0.005727,0.015315,0.003554,-0.007248,0.000749,0.000892,-0.004817,0.011047,0.002982,-0.011714,0.00076,0.003708,-0.008505,-0.00131,-0.004906,0.00419,-0.002048,-0.001676
Gujarati,0.051391,-0.057733,-0.156053,0.110822,-0.080584,0.062193,-0.000223,0.011607,0.033,0.018852,-0.006496,0.000082,-0.002178,0.001741,-0.000679,0.000378,0.00088,-0.000171,-0.001427,-0.003933,0.001429,-0.003314,0.001109,0.001127,-0.002976

3. The founder bottleneck and its medical legacy

The Romani population passed through a severe demographic bottleneck during the migration out of India, and the bottleneck signature is one of the strongest in any European population. IBD segment length distributions, runs of homozygosity, and rare-variant allele frequencies all indicate that the European Romani descend from a founder population of perhaps several hundred individuals. This is small by the standards of human population history and has had striking medical consequences.

A series of rare autosomal recessive disorders have markedly elevated prevalence in Romani populations because of founder mutations that became common in the small founding pool. These include Galactokinase deficiency (Romani type), Congenital cataracts facial dysmorphism neuropathy (CCFDN) syndrome, Hereditary motor and sensory neuropathy Lom type (HMSNL), and others, all caused by founder mutations of South Asian origin that have been preserved in the Romani gene pool despite the subsequent admixture with European populations. The medical-genetic literature on these disorders is, in its own way, a robust independent confirmation of the founder-effect interpretation of Romani population history.

4. The route: Persian, Armenian, and Greek loanword layers

The migration from India to Europe took several centuries, and the linguistic record preserves the geography of the journey with remarkable precision. The Romani language is fundamentally Indo-Aryan in its grammar and core vocabulary, but it carries successive loanword strata that correspond to the major regions through which the migration passed.

The earliest loanword layer is Persian, suggesting a stop of one to two centuries in the Iranian world, probably during the late Sasanian or early Islamic period (roughly 700 to 900 CE). The Romani words for road (drom), village (gav), and many basic items are Persian-derived. The second layer is Armenian, smaller but distinct, suggesting passage through the South Caucasus, possibly during the political turmoil of the Seljuk-Byzantine conflicts in the 11th century. The third and largest non-Indic layer is Byzantine Greek, which dominates the cultural and technical vocabulary of all European Romani dialects. The Greek layer is so substantial that linguists have inferred a sojourn of several centuries in the Byzantine Empire, where the Romani would have been part of the Anatolian and Balkan service population before fanning out into the wider European world after 1300.

5. Entry into the Balkans: the historical record meets the genome

The first unambiguous Byzantine references to a population identifiable as Romani appear in the eleventh century, with mentions of itinerant performers and metalworkers in Constantinople and the surrounding territories. By the 13th century, the population is documented in continental Greek and South Slavic sources. From the late 14th century, Romani communities are recorded across Europe: Wallachia in 1385, Hungary in 1416, German lands in 1417, France in 1419, the Iberian peninsula in 1425, the British Isles in the late 15th century. The European dispersal happened within a window of roughly a hundred years, fanning out from a Balkan base into the broader continent.

The genetic record matches this timing. IBD segment-length analyses across Romani populations of different European countries indicate that the populations diverged from each other approximately 600 to 800 years ago, consistent with a Balkan base ancestor population that fragmented in the late medieval period. The high frequency of shared IBD segments among Romani of distant European countries is itself the signature of this relatively recent dispersal from a common base.

6. The European admixture: where and how

The 70 to 80 percent European autosomal component in modern Romani is not uniform across the diaspora. Different Romani populations carry different European admixture profiles, reflecting the populations among whom they settled. Spanish Roma show strong Iberian autosomal signal, with admixture estimates of 75 to 85 percent Iberian. Romanian and Bulgarian Roma show stronger Balkan and Slavic admixture. Hungarian Roma carry Hungarian and Slavic input. The Welsh Kale, the British Romanichal, the Finnish Kaale, and the various Scandinavian Roma populations each carry the local European autosomal signal of their host population, layered onto the same South Asian founder substrate.

This pattern is the genetic signature of a diaspora population that admixed with surrounding populations in each region while maintaining substantial endogamy, sufficient to preserve the founder bottleneck and the recoverable South Asian signal but not sufficient to prevent significant European gene flow. The degree of admixture correlates loosely with how long the population has been in a given region and how socially integrated or segregated it has been over time. The most isolated Romani sub-populations, such as some of the Vlax Roma of Eastern Europe and certain endogamous Iberian Calo groups, retain the highest South Asian fractions, reaching 35 to 40 percent in some communities.

7. The Y-chromosome and mitochondrial signal

The uniparental markers tell a complementary story to the autosomal data. On the Y-chromosome side, haplogroup H1a1a-M82 is the most diagnostic Romani lineage. It originated in northwest India and reaches 30 to 60 percent in Romani men, depending on the population. Outside the Indian subcontinent, H1a1a is rare. Its presence in European Romani at these frequencies is one of the strongest population-specific Y-chromosome signals known anywhere in Europe.

On the maternal side, mitochondrial haplogroups M5a1 and M18 are the diagnostic Romani lineages. Both are rooted in South Asia and reach 10 to 25 percent in Romani women. The European maternal contribution is also substantial, with the local European mtDNA haplogroups H, U, K, and J appearing at frequencies reflecting the admixture with local populations. The uniparental data are slightly more skewed toward European admixture on the female side than on the male side in some populations, consistent with a pattern of Romani men marrying within the community more strictly than Romani women, or with intermarriage patterns in which non-Romani women joined Romani communities more often than the reverse.

8. The genetic history of the Romani in six phases

9. The myths and the genetic reality

10. Why the Romani case matters for population genetics

The Romani case has become a textbook example in modern population genetics for several reasons. First, it is one of the most precisely dated and most narrowly source-localized migration events in human history, far better resolved than most ancient migrations because it is recent enough that haplotype and IBD signals are still strong. Second, it shows in clean form how a founder bottleneck combined with endogamy can preserve a population-specific genetic signature even after extensive subsequent admixture. Third, it illustrates how genetic, linguistic, and historical data triangulate on the same conclusion when each line of evidence is taken seriously: the language, the loanword strata, the Byzantine and medieval European documents, the Y-chromosome distributions, the mitochondrial DNA, and the autosomal genome all describe the same migration through the same intermediate stops in the same overall time window.

It also stands as a striking example of how the genome can correct centuries of misidentification and stereotyped narrative. The persistent confusion about Romani origins, expressed in everything from medieval Egyptian-origin stories to 19th-century romantic exoticism to 20th-century pseudoscientific racial classifications, has finally been replaced by a clear, evidence-based reconstruction. The Romani are a North Indian diaspora that crossed half of Eurasia in the medieval period and built distinct regional communities in Europe over the past 700 years while preserving, in DNA and in language, the signature of where they came from. The story is, in its own way, one of the great success stories of human population genetics.

11. References

1. Mendizabal, I., Lao, O., Marigorta, U. M., Wollstein, A., Gusmao, L., Ferak, V., et al. (2012). Reconstructing the population history of European Romani from genome-wide data. Current Biology, 22(24), 2342-2349. DOI: 10.1016/j.cub.2012.10.039 Genome-wide Founding event
2. Moorjani, P., Patterson, N., Loh, P.-R., Lipson, M., Kisfali, P., Melegh, B. I., et al. (2013). Reconstructing Roma history from genome-wide data. PLoS ONE, 8(3), e58633. DOI: 10.1371/journal.pone.0058633 Genome-wide Source population
3. Gresham, D., Morar, B., Underhill, P. A., Passarino, G., Lin, A. A., Wise, C., et al. (2001). Origins and divergence of the Roma (Gypsies). American Journal of Human Genetics, 69(6), 1314-1331. DOI: 10.1086/324681 Y-DNA mtDNA
4. Kalaydjieva, L., Calafell, F., Jobling, M. A., Angelicheva, D., de Knijff, P., Rosser, Z. H., et al. (2001). Patterns of inter- and intra-group genetic diversity in the Vlax Roma as revealed by Y chromosome and mitochondrial DNA lineages. European Journal of Human Genetics, 9(2), 97-104. DOI: 10.1038/sj.ejhg.5200597 Vlax Roma
5. Martinez-Cruz, B., Mendizabal, I., Harmant, C., de Pablo, R., Ioana, M., Angelicheva, D., et al. (2016). Origins, admixture and founder lineages in European Roma. European Journal of Human Genetics, 24(6), 937-943. DOI: 10.1038/ejhg.2015.201 Founder lineages
6. Matras, Y. (2002). Romani: A Linguistic Introduction. Cambridge University Press. The reference work on Romani as an Indo-Aryan language and its loanword stratigraphy. Linguistics
7. Fraser, A. (1992). The Gypsies. Blackwell. The standard historical treatment of the European dispersal, with primary source citations for Wallachia 1385, Hungary 1416, German lands 1417, France 1419, Iberia 1425. History
8. Hancock, I. (2002). We Are the Romani People. University of Hertfordshire Press. Comprehensive ethnographic and historical synthesis from a Romani scholar. History Ethnography
9. Kalaydjieva, L., Gresham, D., Calafell, F. (2001). Genetic studies of the Roma (Gypsies): a review. BMC Medical Genetics, 2, 5. DOI: 10.1186/1471-2350-2-5 Founder disorders
10. Davidski, A. (ongoing). Global25 PCA modern population averages. Eurogenes Blog. eurogenes.blogspot.com G25 panel