Kamma and Kapu Caste DNA: History, Origins and What Genetics Shows

Ask a Telugu family where the Kammas or the Kapus "came from" and you will hear many answers. A grandfather in Guntur may talk about Kammanadu and the Vijayanagara generals. An aunt in Rajahmundry may say the Kapus have always been the farmers of the Godavari delta. A cousin in New Jersey may simply want to know why his 23andMe result says "Broadly South Asian" and nothing more.

This guide brings together what historians and population geneticists have actually published about the Kamma and Kapu communities, and about the related Telaga, Balija and Ontari groups. We cover the Kakatiya and Vijayanagara periods, the Nayaka kingdoms and the movement of Telugu families into Tamil Nadu, Telugu naming customs, endogamy, and what DNA studies of Telugu-speaking groups really show. Where the evidence is thin, we say so.

Who are the Kamma and Kapu communities?

Both are large Telugu-speaking communities with roots in Andhra Pradesh and Telangana, and both have spread across South India and, in recent decades, across the world. Neither is a single uniform group. Each contains regional sections, family networks and local traditions, and the boundaries between related groups have shifted over the centuries.

The Kammas

The Kammas are generally described as agriculturists whose name comes from Kammanadu, a region of coastal Andhra lying between the Gundlakamma and Krishna rivers. The region appears as Karma-rashtra in Sanskrit, and the name is found in Ikshvaku-era inscriptions at Jaggayyapeta from around the 3rd century CE, with a Prakrit form, Kammaka-rattha, in later grants. Scholars have offered more than one explanation for the word itself: one links it to the rivers that bounded the region, another, proposed by the scholar Benbabaali, to Buddhist monastic institutions called sanghakammas. The honest summary is that the place name is old, while the community identity built on it is later.

Today Kamma families are concentrated in the Krishna and Guntur region and elsewhere in coastal Andhra, with long-established branches in Telangana and in Tamil Nadu, where Kamma families have lived since the Vijayanagara and Nayaka period.

The Kapus, Telagas, Balijas and Ontaris

In Telugu, kapu literally means cultivator or protector, a meaning noted by the historian Cynthia Talbot. The word has long been used for landholding cultivators in general, which is part of why the history is tangled. As a community name, Kapu today refers mainly to people of coastal Andhra, with a strong concentration in the Godavari and Krishna delta districts.

Kapu, Telaga and Ontari are described in the literature as closely related sections that frequently intermarry and are often counted together, and the terms Kapu and Telaga are frequently used interchangeably. The Balijas, historically associated with trade (the name is usually derived from the Sanskrit vanij, trader), are described as closely related to the Kapus, and the Andhra Pradesh government's Kapu welfare corporation groups Kapu, Telaga, Balija and Ontari together. Some communities with similar names, such as the Munnuru Kapus of Telangana and the Turpu Kapus of north coastal Andhra, are usually described as separate communities.

Neighbouring communities: Reddys and Velamas

No account of Kamma and Kapu history makes sense without the Reddys and Velamas, the other major landholding communities of the Telugu country. Talbot's work, as summarised in standard references, describes several cultivator groups (including those later labelled Reddy, Velama and Kamma) branching into separate named communities in the post-Kakatiya centuries. You can read more about one of them in our guide to Reddy DNA ancestry.

What historians say about origins

Most of what we know about the medieval Telugu country comes from thousands of stone and copper-plate inscriptions. Read carefully, they suggest a society that was more fluid than later caste histories imply.

The Kakatiya period: a fluid, frontier society

The Kakatiya dynasty ruled much of the eastern Deccan from its capital at Orugallu (modern Warangal), with its main period of power running from the 12th century until 1323, when a Delhi Sultanate army under Ulugh Khan (later Muhammad bin Tughluq) took Warangal. The Kakatiya era is remembered for agrarian expansion into the dry uplands, made possible by thousands of irrigation reservoirs, the "tanks", many built by warrior families who served the dynasty. Many of these tanks are still used today.

Talbot's study of Kakatiya inscriptions found that the period's records show surprising flexibility in social rank. She notes that even the varna rank of the Kakatiya rulers themselves is described inconsistently, and the dynasty recruited peasants into its armies, creating new warrior lineages. The title nayaka, meaning a military commander, was open to men of many backgrounds. In other words, the labels people use today were not yet fixed.

Community traditions recorded by colonial ethnographers such as Edgar Thurston tell their own stories about how the Kammas, Velamas and Reddis separated from a common cultivator stock. These are traditions, not documents, and they vary by family and region. They are worth knowing as part of a family's heritage, but they should not be read as literal history.

After 1323: the Musunuri Nayakas

The fall of Warangal brought a period of upheaval. The Musunuri Nayakas, army chiefs who had served the Kakatiyas, united Telugu chiefs and recovered Warangal from the Sultanate, ruling for roughly half a century. Some historians link the Musunuri family to Kamma identity, though, as with many medieval figures, the community labels attached to them are a matter of later interpretation.

Vijayanagara and the Nayaka kingdoms

The Vijayanagara empire (from the 14th to the 17th century) drew heavily on Telugu warriors, administrators and settlers. Historians quoted in standard references describe Kamma commanders and chiefs, among them the Pemmasani, Sayapaneni and Ravella Nayaks, as an important part of Vijayanagara's forces. Balija families served as ministers, generals and provincial governors, and the Nayaka dynasties of Madurai, Thanjavur and Gingee are widely described as being of Balija origin.

This is the period when Telugu-speaking families moved in large numbers into the Tamil country. Scholars such as Velcheru Narayana Rao describe Balija migration into Tamil regions in the 15th and 16th centuries, and after the Battle of Talikota in 1565 many Kamma Nayaks moved to the Madurai and Thanjavur Nayak kingdoms, later settling as agriculturists. That history is why Telugu-speaking Kamma and Balija families (Balijas are known in Tamil Nadu as Kavarai) have been part of Tamil Nadu for centuries. If your family is one of them, our guide to Telugu DNA ancestry covers the wider regional picture.

Colonial era and the deltas

Under British rule, the great irrigation works associated with Arthur Cotton on the Godavari and Krishna deltas transformed farming in coastal Andhra. Kamma and Kapu families, both heavily represented in delta agriculture, were among those affected. Census reports cited by the historian Yamada Keiko recorded Kammas as about 40% of the agricultural population of Krishna district in 1872, rising to 47% by 1921. Caste associations formed in the early 20th century, and in Tamil Nadu Kamma associations grew in the 1920s, particularly around Coimbatore.

Migration did not stop at the Indian border. Kapu and Telaga families were among the Telugu migrants to Burma, Malaysia, Mauritius and Fiji in the colonial era, and Telagas are said to form the largest caste among the Telugu people of Mauritius. After the 1965 US immigration reforms and the IT boom of the 1990s, Kamma, Kapu and other Telugu families became a visible part of the Indian American community. Our Indian American DNA ancestry guide covers what that means for testing.

Names, titles and marriage

The house name comes first

Telugu names traditionally place the family name, the intiperu (literally "house name"), first, followed by the given name. A title is sometimes added at the end, as in the form "Devula Rama Naidu". Abroad, many families reverse the order to fit Western forms, and the intiperu becomes the surname on a passport or green card. Genealogically, the intiperu is often the most useful clue a family has, because it links a lineage to a village or a set of relatives.

Titles are shared, so they are weak clues

Titles such as Naidu (or Nayudu) and Chowdary carry a lot of social meaning, but they are not neat community labels. Naidu, which shares a root with Nayaka, is used by Balija families and also appears among Kamma and other Telugu families. Chowdary is widely associated with Kammas and is traced to revenue-collecting roles under later rulers. Because titles cross community lines, a surname or title is never a reliable stand-in for ancestry, and a genetics paper that labels its sample "Naidu" may not map cleanly onto any single community today.

Endogamy and why it matters for DNA

Like most Indian communities, Kamma and Kapu families have historically tended to marry within the wider community; the Kapu, Telaga and Ontari sections, for instance, are described as frequently intermarrying with one another. Many Indian communities also have rules about which family lines may not marry each other. For genetics this matters a great deal. When a community marries within itself for many generations, its members come to share more stretches of DNA with each other than with outsiders, and small differences in allele frequencies build up. That is what lets researchers tell closely related neighbouring groups apart at all. It is also what makes community reference groups useful. For the bigger picture, see what genetics shows about caste and endogamy and how gotras compare with DNA lineages.

What population genetics shows for Telugu-speaking groups

Andhra Pradesh has been unusually well represented in Indian genetics, thanks partly to long-running research collaborations involving Andhra University. Here is what the major studies found.

Reich and colleagues, 2009: two ancestral populations

The landmark 2009 Nature paper by Reich and colleagues showed that most Indian groups descend from a mixture of two ancient populations: Ancestral North Indians (ANI), related to people of the Middle East, Central Asia and Europe, and Ancestral South Indians (ASI), not closely related to any group outside the subcontinent. ANI ancestry ranged from about 39% to 71% across Indian groups. Seven of the 25 Indian groups in the study came from Andhra Pradesh: groups labelled Vysya, Naidu, Velama, Madiga, Mala, Kamsali and the Chenchu tribe. The Andhra caste groups spanned a wide part of that range, from about 39% to about 55% ANI, which is a reminder of how much variation exists inside a single state. The paper also identified strong founder events in several groups, including the Vysya, dated to more than 100 generations ago. Kammas and Kapus were not among the sampled groups. Our explainer on ANI and ASI ancestry covers the model in detail.

Moorjani and colleagues, 2013: when the mixing happened

Using genome-wide data from 73 groups, Moorjani and colleagues dated the mixing of ANI and ASI to roughly 1,900 to 4,200 years ago. For the Andhra groups in their table, the estimates ran from about 2,500 years ago (Velama) to about 4,200 years ago (Vysya). Dates were typically more recent in Indo-European speaking groups than in Dravidian speaking ones. The authors describe India as moving from a region where mixture was common to one where even closely related groups rarely mixed, because of a shift to endogamy.

Basu and colleagues, 2016: endogamy about 70 generations ago

A 2016 PNAS study by Basu, Sarkar-Roy and Majumder analysed 18 mainland and 2 island populations and found four major ancestries in mainland India rather than two. They estimated that widespread mixing was replaced by endogamy about 70 generations ago, most sharply among upper castes and Indo-European speakers, and suggested this coincides with the Gupta period. Their sample did not include any Telugu-speaking group, so it describes the national picture rather than Andhra specifically.

The deep layers: Indus, Steppe and AASI

Ancient DNA has since refined the ANI and ASI model. The 2019 Science paper by Narasimhan and colleagues, based on 523 ancient genomes, found that after the decline of the Indus Valley Civilisation its people mixed with groups to the southeast to form ASI, "whose direct descendants live in southern India", while Steppe-related ancestry arrived in the north from around 4,000 years ago. A 2019 Cell paper by Shinde and colleagues showed that an ancient genome from Rakhigarhi fitted as a mixture of ancestry related to ancient Iranians (from a lineage that split off before farming began on the Iranian plateau) and a component related to Southeast Asian hunter-gatherers, a stand-in for the ancient AASI lineage, with little if any Steppe ancestry. Telugu communities, like other South Indian groups, carry these layers in different proportions. Our guide to Steppe ancestry in India explains the later layer.

Studies that actually sampled Kammas or Kapus

Here is the part most websites skip. When we searched the peer-reviewed literature, we found only a handful of studies that sampled Kammas or Kapus by name, and most of them used older, lower-resolution markers. Anyone quoting precise "Kamma ANI percentages" or "Kapu haplogroup frequencies" should be able to tell you which paper they come from.

Kammas: Y chromosome lineages shared across boundaries

In 2001 Ramana and colleagues, writing in the European Journal of Human Genetics, typed 27 Y chromosome markers in 204 men from Andhra Pradesh: Vizag Brahmins, Peruru Brahmins, Kammas, three tribal groups and the Siddis. They found paternal haplotypes shared across caste and tribe boundaries. In their analysis the difference between the caste and tribal categories as a whole was not statistically significant, while differences between individual populations within those categories were. Their conclusion was that gene flow among these South Indian populations was "not as negligible" as earlier studies had suggested. For a Kamma reader, the takeaway is simple: paternal lineages do not respect community boundaries neatly.

Kapus: low-coverage genomes from Andhra Pradesh

A 2017 BMC Genomics study by Rustagi and colleagues sequenced 185 people from eight Andhra Pradesh populations at very low coverage, including 37 Kapu individuals. On the main axis of genetic variation, the Kapu and Yadava samples were indistinguishable from the 1000 Genomes Project samples of Telugu people living in the UK, Bengalis and Sri Lankan Tamils, while being distinguishable from some other Andhra groups in the study. Their mitochondrial data showed that the caste populations studied, with one exception, had at least 30% non-M maternal lineages, consistent with more gene flow into caste populations than into isolated tribal groups.

An earlier 2008 study in BMC Genetics by Watkins and colleagues compared Tamil caste groups with Andhra samples that included Kapus. Its most useful finding for readers is this: genetic differentiation among the Tamil castes studied was low, reflecting "a largely common origin", and the variation between castes within one region was larger than the variation between the Tamil and Andhra samples as wholes. Put simply, neighbouring communities differ only a little, and region matters as much as community.

Founder events in Telugu groups

The largest study of community structure in South Asia, published in Nature Genetics in 2017 by Nakatsuka and colleagues, assembled data from over 2,800 people in over 260 groups. It found 81 groups descended from founder events stronger than those of Finns and Ashkenazi Jews, of which 14 have census sizes over a million. Five of those 14 were Telugu groups: samples labelled Pattapu Kapu, Vadde and Kshatriya Agnikula from Andhra Pradesh, and Reddy and Vysya from Telangana. The paper's point was medical, not social: strong founder events can raise the frequency of particular recessive conditions, so community-aware genetics could help prevent disease. A founder event says nothing about status or worth. It simply means that a community's present-day members descend from a relatively small set of ancestors at some point in the past.

Two cautions. The sample labelled Pattapu Kapu represents one sampled subgroup, not every Kapu, Telaga or Balija family. And we found no peer-reviewed genome-wide study labelled Balija or Ontari, so anything stated about their genetics should be treated as unverified.

Kamma vs Kapu: what genetics can and cannot say

People search for "Kamma vs Kapu" expecting a genetic verdict. The evidence does not support one. Both are Dravidian-speaking communities of the same region with a long shared history, and the published data suggest that, like neighbouring communities across South India, they are genetically close. The table below summarises what is known.

Community	Main regions	Traditional occupation (per historians)	What published genetics can say
Kamma	Krishna and Guntur (Kammanadu), coastal Andhra; branches in Telangana and Tamil Nadu	Agriculture; military and administrative service under Vijayanagara and later rulers	A Y chromosome study (Ramana 2001) found paternal lineages shared across caste and tribe boundaries. We found no published genome-wide Kamma estimate.
Kapu / Telaga	Godavari and Krishna deltas, coastal Andhra	Cultivation (kapu means cultivator or protector); many Nayaka-era commanders were Telagas	Low-coverage genomes of 37 Kapus (Rustagi 2017) clustered with Telugu samples from the UK. A Pattapu Kapu sample showed a strong founder event (Nakatsuka 2017).
Balija	Andhra Pradesh, including Rayalaseema; Tamil Nadu (Kavarai); Karnataka	Trade, with warrior-merchant roles; Nayaka rulers of Madurai, Thanjavur and Gingee	We found no peer-reviewed genome-wide study labelled Balija.
Ontari	Coastal Andhra, counted with Kapus	Described as a warrior section in medieval Andhra	No published genetic data found.
Velama	Telangana and Andhra Pradesh (the Reich 2009 sample came from Mahbubnagar)	Cultivator lineages that became a named community after the Kakatiya period (Talbot)	Sampled in Reich 2009 and Moorjani 2013; ANI-ASI mixture dated to roughly 2,500 years ago.
Reddy	Telangana and Andhra Pradesh	Cultivator lineages (Panta and Pakanati Kapus) that became a named community after the Kakatiya period (Talbot)	A Telangana Reddy sample showed a founder event stronger than that of Finns (Nakatsuka 2017).

The table describes populations, not people. It does not rank the communities, and none of these findings implies that any group is "purer", older or more distinguished than another. Every one of these communities descends from the same deep ancestral sources, mixed in somewhat different proportions and shaped by different marriage histories.

Why your result might surprise you

Several things can make a result look different from what the family expects. Every person has thousands of ancestors, and a few generations of marriage outside the community can shift similarity scores. Communities that are genetically close, such as neighbouring landholding groups in the same district, can swap places at the top of a similarity list. Regional ancestry often shows up more strongly than community, which fits the Watkins finding above. And reference groups are samples, not entire communities. If you want the long answer, read can a DNA test tell your caste?

What a Helixline test shows for Telugu families

Helixline's analysis compares your DNA with a reference panel of 7,600+ global reference samples and 140+ South Asian community reference groups. Telugu groups in the panel include Kamma, Kapu, Balija, Nayak/Balija, Naidu, Reddy, Velama, Telugu Brahmin, Arya Vysya/Komati, Madiga and Mala, along with the Chenchu and Southern Gond groups. Results are similarity comparisons with those groups, presented alongside your state-level breakdown.

In India: the Decode kit

• Origins (₹6,999): state-level ancestry breakdown, community similarity comparisons, ANI, ASI, Steppe and AASI composition, and your Y-DNA and mtDNA haplogroups.
• Decode (₹12,999): everything in Origins, plus health, traits and pharmacogenomics reports and a Health Coach consultation. Choose it when you want more than ancestry.
• Infinite (₹17,999): everything in Decode, plus a 137-marker Full Body Blood Test and a doctor consultation.

The kit uses a saliva sample and is collected from your door at no charge. Kit in 3-4 days, results in 2-3 weeks from the moment the sample reaches the lab, an ISO 15189 certified Indian lab partner, and your data stays in India. Order at helixline.in/buy.

Telugu Americans and families abroad: upload your file

If you have already tested with 23andMe, AncestryDNA, MyHeritage or FamilyTreeDNA, you do not need a new kit. Upload your raw data file at helixline.in/upload and get results within 1-2 days.

Plan	Price (one-time)	What you get
Upload Ancestry	$25 · £19 · €23 · CA$34 · A$38 · AED 92 · S$34 · ₹1,999	Community similarity comparisons, state-level breakdown, haplogroups where your file has the markers
Upload Complete	$50 · £39 · €46 · CA$68 · A$76 · AED 184 · S$68 · ₹3,999	Everything above plus wellness and health reports

Gifting a kit to parents in India

Many Telugu families abroad find that the most interesting results come from their parents or grandparents, whose ancestry sits a generation closer to the village and the intiperu. You can buy a kit from abroad for family in India at helixline.in/buy (choose "Someone in India" at checkout) or see the gift page. Kits ship within India as standard, and delivery to other countries can be arranged on request with a quote through our international page.

Reading results as a family

Test more than one relative if you can, ideally from both sides of the family. Siblings share about half their DNA, so two siblings can see slightly different similarity lists, and that is normal. Keep the oral history and the paper records alongside the DNA: village names, the intiperu, temple records and old letters answer questions about named ancestors that genetics cannot.

Sources and further reading

• Reich D, et al. Reconstructing Indian population history. Nature, 2009. pmc.ncbi.nlm.nih.gov/articles/PMC2842210
• Moorjani P, et al. Genetic evidence for recent population mixture in India. American Journal of Human Genetics, 2013. pmc.ncbi.nlm.nih.gov/articles/PMC3769933
• Basu A, Sarkar-Roy N, Majumder PP. Genomic reconstruction of the history of extant populations of India reveals five distinct ancestral components and a complex structure. PNAS, 2016. pmc.ncbi.nlm.nih.gov/articles/PMC4760789
• Narasimhan VM, et al. The formation of human populations in South and Central Asia. Science, 2019. pmc.ncbi.nlm.nih.gov/articles/PMC6822619
• Shinde V, et al. An ancient Harappan genome lacks ancestry from Steppe pastoralists or Iranian farmers. Cell, 2019. pmc.ncbi.nlm.nih.gov/articles/PMC6800651
• Ramana GV, et al. Y-chromosome SNP haplotypes suggest evidence of gene flow among caste, tribe, and the migrant Siddi populations of Andhra Pradesh, South India. European Journal of Human Genetics, 2001. pubmed.ncbi.nlm.nih.gov/11571559
• Rustagi N, et al. Extremely low-coverage whole genome sequencing in South Asians captures population genomics information. BMC Genomics, 2017. pmc.ncbi.nlm.nih.gov/articles/PMC5440948
• Watkins WS, et al. Genetic variation in South Indian castes: evidence from Y-chromosome, mitochondrial, and autosomal polymorphisms. BMC Genetics, 2008. pmc.ncbi.nlm.nih.gov/articles/PMC2621241
• Nakatsuka N, et al. The promise of discovering population-specific disease-associated genes in South Asia. Nature Genetics, 2017. pmc.ncbi.nlm.nih.gov/articles/PMC5675555
• Background reading on history (with references to Talbot, Stein, Benbabaali and others): Kakatiya dynasty, Kamma, Kapu, Balija and Telugu names.

Frequently Asked Questions

What is the origin of the Kamma caste?

Historians link the Kamma name to Kammanadu, a region of coastal Andhra between the Gundlakamma and Krishna rivers that appears in inscriptions from around the 3rd century CE. The community is described as agriculturist, and its identity took clearer shape after the Kakatiya period, when Kamma commanders served the Vijayanagara empire and some later moved into the Nayaka kingdoms of Tamil Nadu. Family and community traditions about earlier origins vary and should be read as heritage rather than documented history.

Who are the Kapu, Telaga, Balija and Ontari communities?

Kapu means cultivator or protector in Telugu. Kapu, Telaga and Ontari are described as closely related sections of coastal Andhra that frequently intermarry and are often counted together, and the terms Kapu and Telaga are often used interchangeably. Balijas, historically associated with trade and with the Nayaka rulers of Madurai, Thanjavur and Gingee, are described as closely related, and the Andhra Pradesh government groups all four together for its Kapu welfare corporation.

Is there a genetic difference between Kamma and Kapu?

Published studies give no basis for a genetic verdict between them. Both are Dravidian-speaking communities of the same region, and research on South Indian castes finds low genetic differentiation between neighbouring communities, with variation between castes in one region larger than the variation between regions. Very few studies have sampled Kammas or Kapus directly, and none of the evidence supports ranking one community above another.

Can a DNA test tell if I am Kamma or Kapu?

No. DNA cannot identify or prove a caste. Caste is a social identity passed on through family and community, not a genetic marker. A test can compare your DNA with reference groups, such as the Kamma and Kapu groups in Helixline's panel, and report how similar you are to each. That is a statement about shared ancestry, never proof of membership.

What do genetic studies show about Telugu communities?

Reich and colleagues (Nature, 2009) found that Indian groups, including seven from Andhra Pradesh, are mixtures of Ancestral North Indian and Ancestral South Indian populations, with Andhra caste groups spanning roughly 39% to 55% ANI. Moorjani and colleagues (2013) dated that mixing to roughly 1,900 to 4,200 years ago. Rustagi and colleagues (2017) sequenced 37 Kapu individuals, whose genomes clustered with Telugu samples from the UK, and Nakatsuka and colleagues (2017) found strong founder events in several large Telugu groups.

How can Telugu families in the USA test their ancestry?

If you have already tested with 23andMe, AncestryDNA, MyHeritage or FamilyTreeDNA, upload your raw data to Helixline from $25 (or the equivalent in 8 currencies) for similarity comparisons against 140+ South Asian community reference groups, including Kamma, Kapu, Balija, Reddy and Velama groups, with results within 1-2 days. You can also buy a kit for parents in India at helixline.in/buy by choosing Someone in India at checkout.

Curious where your own DNA sits among Telugu reference groups? In India, order the Decode kit. Abroad, upload your 23andMe or AncestryDNA file from $25.