Sri Lankan Tamil DNA Ancestry: What Your 23andMe or AncestryDNA Result Means

If your family is Sri Lankan Tamil and you tested with 23andMe or AncestryDNA in Toronto, London, Sydney, New Jersey or Zurich, your result probably landed on one of a few labels: Sri Lankan, Southern Indian, a combined South Indian and Sri Lankan group, or Broadly South Asian. The wording depends on the company and on when you last looked, because vendors rename their regions.

None of those labels is wrong. They are coarse. This guide sets out what published genetics says about Sri Lankan Tamil ancestry, how it relates to Tamils in India, the Sinhalese and the island's Indigenous Adivasi (Vedda), which lineages are common on the island, and what your raw data file can and cannot show.

Who the Sri Lankan Tamils are, in population terms

Sri Lanka's 2012 census counted Sri Lankan Tamils at 11.2% of the population, alongside the Sinhalese at 74.9%, Muslim communities at 9.2% and Indian Tamils at 4.2%, as quoted in the 2025 whole-genome study by Urban Aragon and colleagues, the source of most numbers below.

The island has two Tamil communities with different histories. Sri Lankan Tamils, historically concentrated in the north and east, have deep roots there: linguistic evidence suggests Tamil settlement in the north began at roughly the same time as Sinhala settlement, in the first millennium BCE. Indian Tamils, often called Hill Country or Malaiyaha Tamils, descend largely from workers brought from South India in the British colonial period.

Much of the diaspora in Canada, the UK, Australia and Europe left during the civil war that began in the 1980s. DNA neither adds to that history nor takes from it; it describes the far older history underneath.

What the research shows about Sri Lankan Tamil ancestry

The South Asian cline, in brief

In 2009, Reich and colleagues showed that most groups in India are mixtures of two ancestral populations, Ancestral North Indians (ANI) and Ancestral South Indians (ASI), in varying proportions. Moorjani and colleagues dated that mixing to roughly 1,900 to 4,200 years ago; the 2025 Sri Lankan study put the arrival of West Eurasian-related ancestry on the island at about 3,000 years ago, an average over more than one event.

The 2019 ancient DNA study by Narasimhan and colleagues went deeper. ASI was itself a mixture: people related to the Indus Periphery, carrying Iranian-farmer-related ancestry, mixed with groups high in AASI, an ancient South Asian hunter-gatherer ancestry with no unmixed living descendants, represented only distantly by present-day Andamanese. ANI formed when steppe pastoralist ancestry from Central Asia mixed into the region, and both took shape in the second millennium BCE. See our explainers on AASI and ANI and ASI.

Where Sri Lankan Tamils sit on that cline

The 2025 Current Biology study sequenced 35 Sinhalese and 19 Adivasi genomes and added 35 Sri Lankan Tamils sampled in the UK by the 1000 Genomes Project (population code STU). Using Narasimhan's three-source model, with Andamanese as a distant stand-in for AASI, it estimated:

Population (as sampled)	AASI-related ancestry	What else the study reports
Sri Lankan Tamils, UK (1000 Genomes STU)	45.4%	More Iranian-farmer-related ancestry than the Adivasi; steppe under 10%
Sinhalese, Sri Lanka	46.9%	Very similar profile to the Sri Lankan Tamils
Interior Adivasi (Rathugala)	52.5%	Strong recent founder event and small population size
Coastal Adivasi (Echchalampaththu)	53.7%	Milder founder event than the Interior clan
Urban South Indian groups (Telugu in the UK, Yadava, Urban Bangalore, Urban Chennai)	about 40 to 50%	The range the Sri Lankan Tamils and Sinhalese fall into
Mala, Madiga and Kapu (South India)	over 50%	The range the Adivasi fall into

These are model estimates, not fixed properties of a people. A different reference set or AASI proxy gives different numbers for the same genomes, which is why online calculators often report a much higher hunter-gatherer share. Compare like with like.

Tamils and Sinhalese: closer than their languages suggest

Sinhala is Indo-European and Tamil is Dravidian, yet the two communities differ far less genetically than their languages. A 2023 iScience study by Singh and colleagues, genotyping Sinhalese and Sri Lankan Tamils alongside the UK-sampled STU panel, found a tight cluster of Sinhalese and Tamil populations, suggesting strong gene flow beyond the boundary of ethnicity and language, and found Sri Lankan groups more homogeneous than others in South Asia.

There is a faint distinction. Haplotype analysis preserved a trace of North Indian affiliation in the Sinhalese, and shared short DNA segments pointed to deep common roots with the Maratha. The 2025 study likewise grouped the Sinhalese and Sri Lankan Tamils together, with the two Adivasi clans as a separate sub-cluster. In practice, no consumer test can reliably tell a Tamil grandparent from a Sinhalese one.

Links to Tamils and other communities in India

Beyond the island, all the Sri Lankan groups in the 2025 study shared the most genetic drift with southern Indian populations such as the Mala and Madiga, and the authors suggest the Mala may be closely related to the sources behind the Sinhalese, Sri Lankan Tamils and Adivasi. Mala and Madiga are Telugu-speaking: genetic similarity does not follow language or state lines, so expect a Sri Lankan Tamil genome to resemble a spread of South Indian groups. For the mainland picture, see Tamil DNA ancestry and Dravidian ancestry genetics.

Maternal lineages agree. Ranaweera and colleagues sequenced mitochondrial DNA from 271 people in five groups (Vedda, Up-country and Low-country Sinhalese, Sri Lankan Tamils and Indian Tamils) and found the Sinhalese and Tamils aligned with Indian subcontinental populations, while the Vedda stood apart. Singh and colleagues found more West Eurasian maternal lineages in Sri Lankan groups than in South India, except among Indian Tamils, consistent with their recent colonial-era migration, and the Moors.

The Vedda, or Adivasi

Modern humans were in Sri Lanka by around 40,000 years ago, and the Indigenous Vedda are proposed to descend in part from those early inhabitants. The 2025 study uses the more respectful name Adivasi, meaning original inhabitant, and sampled two clans: the Interior Adivasi of Rathugala, surrounded by Sinhalese communities, and the Coastal Adivasi of Echchalampaththu, next to Sri Lankan Tamil communities.

Both clans carry more AASI-related and less Iranian-farmer-related ancestry than the Sinhalese or Sri Lankan Tamils, and both show founder events, the Interior clan's the stronger, about 210 years ago. Yet they remain genetically close to their neighbours: the authors conclude that the ancestors of these clans admixed heavily with later migrants such as the ancestors of today's Sinhalese and Sri Lankan Tamils, and intermarriage is not uncommon today.

Some Sri Lankan Tamil families may therefore carry Adivasi-related ancestry, but a small share in one person is very hard to detect, and no responsible service should promise to label it.

Haplogroups common in Sri Lanka

The Y chromosome (in men) and mitochondrial DNA each trace one unbroken line, father to son and mother to child. Published counts come mainly from small Sinhalese and Adivasi samples, so read them as the island's broad picture, not Sri Lankan Tamil frequencies.

• mtDNA M: carried by 22 of 35 Sinhalese in the 2025 study; M is the commonest maternal family in South Asia (what haplogroup M means).
• mtDNA R30b2a: the most frequent Adivasi lineage, in 4 of 5 Interior and 6 of 14 Coastal individuals.
• mtDNA U: U7 and U1 are the commonest West Eurasian maternal lineages reported in Sri Lanka; U2c1a appears in the Adivasi (haplogroup U in South Asia).
• Y-DNA H (H1a, H1b, H3b): found in 7 of 28 Sinhalese men, and in both Adivasi clans.
• Y-DNA J2a: 5 of 28 Sinhalese men (J2 in India).
• Y-DNA R2a: 3 of 28 Sinhalese men, plus Adivasi men in both clans.
• Y-DNA L1a1: 2 of 28 Sinhalese men, a lineage widespread in West and South Asia.

For mainland comparisons, see Y-DNA haplogroup frequencies in India and Indian maternal haplogroups.

What your 23andMe or AncestryDNA result can and cannot tell you

A consumer result compares your DNA with the company's reference panel, so it can only be as specific as that panel, and global panels hold far fewer South Asian samples than European ones. That is why Broadly South Asian exists (why 23andMe says Broadly South Asian).

What it can tell you

• That your ancestry is South Asian and largely southern. For most Sri Lankan Tamils that is the expected answer, and accurate at that scale.
• Who your relatives are. DNA matching is often the most valuable part for families separated by migration, and an upload service cannot replace it.
• A broad haplogroup, where your vendor reports one.

What it cannot tell you

• Sri Lankan Tamil versus Tamil Nadu ancestry, with confidence. The populations are closely related and thinly sampled, so a split between the two labels does not measure where your family lived.
• Tamil versus Sinhalese ancestry. As above, the research shows a tight cluster.
• Jaffna versus Batticaloa, or any district. We know of no reference data that resolves the island at that level, and we do not claim to.
• Your deep ancestry layers. Neither company reports AASI, Iranian-farmer-related or steppe proportions.
• Caste, religion or identity. These are social facts, not genetic ones.

What re-analysing your raw file adds

The raw file behind your result holds several hundred thousand measured positions. The measurement is fixed; the interpretation can be redone. Uploaded to Helixline, it is compared against South Asian reference populations, including the 140+ community clusters described on this site, and your report shows:

• Your closest reference groups, ranked by genetic similarity. For a Sri Lankan Tamil file, expect matches spread across South Indian and Sri Lankan references: a statement about similarity, not a verdict on who you are.
• AASI, Iranian-farmer-related and steppe proportions, to set beside published figures, with the model caveat above.
• Y-DNA and mtDNA haplogroups, where your file carries those markers.

Comparing services? See the best DNA upload sites for South Asian ancestry.

How to download your raw data and upload it

Step 1: download the file from your testing company

Menus change, but the route is similar everywhere: find the raw data download in your DNA or account settings, confirm by email if asked, and save the file.

• 23andMe: in your account settings, request the raw data download; it arrives as a .zip. If you have not signed in for a while, check the account is still active. See the 23andMe raw data format.
• AncestryDNA: in the DNA section, open your test settings, choose Download Raw DNA Data and confirm by email (full walkthrough).
• MyHeritage: under DNA, choose Manage DNA kits, then your kit's actions and Download raw DNA data (MyHeritage guide).
• FamilyTreeDNA and LivingDNA: take the autosomal raw data file from your account's download section.

Step 2: check the file, and do not edit it

Keep the .txt or .zip exactly as downloaded, and never save it from a spreadsheet, which can silently corrupt it. Our raw DNA file check tells you whether a download is complete.

Step 3: upload to Helixline

1. Go to helixline.in/upload and click through to app.helixline.in, where the upload and payment take place.
2. Upload your unmodified .txt or .zip file. We accept raw data from 23andMe (v3, v4 and v5 chips), AncestryDNA, MyHeritage, FamilyTreeDNA and LivingDNA.
3. Choose the Ancestry report at ₹1,999 / $25, or the Complete report (ancestry and health) at ₹3,999 / $50.
4. Pay. Results normally arrive within 24 hours of payment and a valid file.

Your data is stored encrypted in India, and uploading works the same from Canada, the UK, Australia, the US or Europe, with nothing to ship. More in our diaspora DNA guide.

A note on DNA and identity

Being Sri Lankan Tamil is a matter of language, history, family and community. A genome does not decide it, and a percentage does not confirm or cancel it. Read any ancestry result, ours included, as genetic similarity to sampled groups, not a statement about caste, religion or belonging.

Frequently Asked Questions

Is Sri Lankan Tamil DNA different from Tamil Nadu Tamil DNA?

Only slightly. Sri Lankan Tamils sit on the same genetic cline as South Indians, with AASI, Iranian-farmer-related and steppe proportions in the range of urban South Indian groups. Research finds Sri Lankan groups closest to southern Indian populations and more homogeneous than mainland groups, but the differences are too subtle for consumer tests to separate reliably.

Why does 23andMe say Broadly South Asian for Sri Lankan Tamils?

Because the result reflects the reference panel, not your DNA. When the company's South Asian references cannot place a stretch of your genome confidently, it falls back to a broad label. A panel with more South Asian populations can usually say more from the same file.

Can DNA tell Sri Lankan Tamil and Sinhalese ancestry apart?

Not reliably. Published studies find a tight genetic cluster of Sinhalese and Sri Lankan Tamils, reflecting long gene flow across language and ethnicity. Research groups have detected faint differences, such as a trace of North Indian affinity in the Sinhalese, but no consumer test can assign an individual to one community.

Do Sri Lankan Tamils have Vedda ancestry?

Some may. A 2025 whole-genome study concluded that the ancestors of the Adivasi (Vedda) mixed heavily with later migrants, including the ancestors of Sri Lankan Tamils and Sinhalese, and intermarriage continues today. Because the groups are genetically close, a small Adivasi-related share in one person is very hard to detect.

Can I upload my 23andMe or AncestryDNA file from Canada, the UK or Australia?

Yes. Uploading involves no shipping, so it works from anywhere. Download the raw file, go to helixline.in/upload and click through to app.helixline.in to upload and pay. The Ancestry report costs $25 and results normally arrive within 24 hours of payment and a valid file.

Conclusion

A Sri Lankan, Southern Indian or Broadly South Asian label is a correct but blunt summary. Underneath it, published genetics describes a population built from AASI, Iranian-farmer-related and a little steppe ancestry, closely related to the Sinhalese, sharing deep links with South Indian communities, and in places mixed with the Adivasi. The file you already have can show where you fall in that picture, read as similarity rather than identity.

Read next: Tamil DNA: Dravidian roots and ASI ancestry.