Sephardic Jews | Conversos | Iberia | Latin America
Can DNA Prove a Converso Ancestor?
Many Spanish, Portuguese and Latin American families tell of a hidden Jewish ancestor. Large studies find real Sephardic traces across Latin America, but our own tests show what a single DNA test can, and cannot, say about one family.
Hero image: AI generated illustration of Sephardic families leaving an Andalusian harbour in 1492, a man carrying a Torah scroll, horseshoe arches behind them and caravels leaving for the Ottoman Empire. An artistic reconstruction, not a historical document.
"My grandmother lit candles on Friday night and never knew why." Stories like this are told in thousands of families in Spain, Portugal, Mexico, Colombia, Brazil and beyond. They point to the conversos, the Jews of Iberia who were forced to convert to Christianity, and to the descendants who kept a few customs for centuries. DNA tests promise to check such stories. Here is what the science really shows, and what our own Global25 tests can and cannot detect.
From Sefarad to the conversos
For more than a thousand years, Iberia, called Sefarad in Hebrew, was home to one of the largest Jewish communities in the world. The pogroms of 1391 forced tens of thousands of Jews to convert. The Spanish Inquisition, founded in 1478, pursued the converts suspected of practising Judaism in secret. In 1492 the Alhambra Decree expelled the Jews who refused baptism from Spain; many went to Portugal, where the whole community was forcibly converted in 1497.
The exiles founded the Sephardic communities of the Ottoman Empire (Thessaloniki, Istanbul, Izmir, Sofia), of Morocco, and later of Amsterdam, London and the Caribbean. Those who stayed became "New Christians". Despite official bans, many reached the Americas, and the Inquisition followed them: its tribunals in Mexico City, Lima and Cartagena tried hundreds of people accused of secret Judaism. In Portugal, the crypto-Jews of Belmonte kept their traditions in secret until the twentieth century.
What Sephardic DNA looks like
In 2026, a team from the Autonomous University of Barcelona published the first genomes of a medieval Iberian Jewish community: victims of the 1348 pogrom buried at Roquetes in Tàrrega, Catalonia. In one model, about two thirds of their ancestry came from a source close to the Bronze Age people of Canaan, and the rest from medieval Iberians. Their background endogamy was lower than in the medieval Jewish communities of Erfurt and Norwich.
Studies of paternal lines point the same way. In 2008, Susan Adams and colleagues estimated that about a fifth of Iberian men carry Y-chromosome lineages attributed to Sephardic Jewish origins, while noting that part of this signal may come from older movements across the Mediterranean.
Today's Sephardic Jews of Turkey, Bulgaria, Greece and Morocco show the same kind of mix: a Levantine base with a large share of Southern European ancestry. This is exactly what makes converso ancestry hard to measure. In Global25, Sephardic Jews sit close to Southern Italians, Greeks and Levantines, so a model can easily replace a Sephardic ancestor with a Sicilian one, or the other way round.
What the large study found in Latin America
In 2018, Juan Camilo Chacón-Duque and colleagues analysed 6,589 people from Brazil, Chile, Colombia, Mexico and Peru with haplotype-based methods, which compare long stretches of DNA rather than averages. They found ancestry related to Sephardic and Eastern Mediterranean sources across the whole region, most likely brought by conversos during the colonial period. About a quarter of the participants carried more than 5% of it, but the averages by country were small, from about 1% in Brazil to 4% in Chile.
In other words, converso ancestry is real and widespread in Latin America, but it is usually a thin layer, spread over many lines of a family tree, not a single ancestor that a test can point to.
Our test: how far can Global25 go?
We modelled people with the regional populations of Spain and Portugal, other Southern Europeans, North Africans, and five Sephardic references: the Sephardic Jews of Turkey, Bulgaria and Greece, Moroccan Jews, and the medieval Jews of Tàrrega. For Latin Americans we added Native American and West African sources. Then we asked a simple question: what "Sephardic" share does the model give to people who have no known Jewish ancestor, and what does it give to people who certainly do?
| Who | "Sephardic" share found by the model |
|---|---|
| Ordinary Spaniards and Portuguese (319 individuals) | average 4.8%; 1 in 4 above 8%, 1 in 10 above 15%, 1 in 20 above 20%, maximum 27% |
| Andalusians / Portuguese / Galicians | average 7.8% / 7.8% / 5.6% |
| Simulated: 1 Sephardic great-great-grandparent (6%) | 0 to 29%, average 11% |
| Simulated: 1 Sephardic great-grandparent (12.5%) | 0 to 35%, average 14% |
| Simulated: 1 Sephardic grandparent (25%) | 0 to 43%, average 21% |
| Simulated: 1 Sephardic parent (50%) | 15 to 62%, average 40% |
The result is sobering. An ordinary Spaniard can get more "Sephardic" ancestry in this model than a person who really has a Sephardic grandparent. Only a Sephardic parent stands out clearly. The reason is the one above: Iberians themselves carry Eastern Mediterranean and North African ancestry from Antiquity and the Middle Ages, and Sephardic DNA overlaps with it.
| Population average | "Sephardic" share |
|---|---|
| Mexican (Euro-Mestizo) | 13.8% |
| Colombian (Euro-Mestizo) | 11.8% |
| Euro-Argentine | 8.3% |
| Colombian (Mestizo) | 7.2% |
| Mexican (Mestizo) | 6.0% |
| Cuban, Puerto Rican, Peruvian (Mestizo) | 4.5 to 4.7% |
| Brazilian | 3.8% |
| Dominican | 2.0% |
What this means for your family story
A converso ancestor from the 1500s or 1600s cannot be proven or ruled out by an ancestry percentage. Fifteen generations back, each person has thousands of ancestors, and many of them left no DNA at all in their descendants. Converso ancestry in Latin America is usually a thin layer spread over many lines, and it looks like the Mediterranean ancestry that most Iberians already carry.
What DNA can do is more modest, and still useful:
- Detect recent Sephardic ancestry. A Sephardic parent, and sometimes a grandparent, does stand out from the normal range of your country.
- Compare you with your own country. A "Sephardic" result only means something when it is clearly above what ordinary people from the same region get.
- Find relatives. DNA matches with known Sephardic families (Turkey, Greece, Bulgaria, Morocco, Amsterdam, the Caribbean) are stronger evidence than any percentage. Haplotype-based tests like those of MyHeritage and 23andMe are better at this than Global25.
- Read the paternal and maternal lines with care. Some Y-DNA and mtDNA branches are frequent among Sephardim, but most of them are also common among non-Jewish Iberians. Only a precise sub-branch shared with documented Sephardic families is a real clue.
- Go back to the records. Inquisition trials, notarial archives and parish books are often the only way to reach the 1500s. A surname alone proves nothing: most "converso surnames" lists include names carried by millions of Old Christians.
Test the legend in your family
Our Legend Check (7 EUR) tests the family stories people ask us about most, including a hidden Jewish ancestor, a Moorish or Carthaginian ancestor, a Romani ancestor and an Amerindian ancestor. Each legend gets an honest verdict, and the report tells you when the data cannot decide, as in the cases above. For the deep regional story of Iberian ancestry, see our Hesperia report, and for Latin American roots, Latino Roots.
References
- Chacón-Duque JC, et al. (2018). Latin Americans show wide-spread Converso ancestry and imprint of local Native ancestry on physical appearance. Nature Communications 9: 5388. doi.org/10.1038/s41467-018-07748-z
- Genomic study of the medieval Jewish community of Tàrrega (Roquetes necropolis), Autonomous University of Barcelona team led by Cristina Santos (2026). Genes 17(3): 358. mdpi.com/2073-4425/17/3/358
- Adams SM, et al. (2008). The genetic legacy of religious diversity and intolerance: paternal lineages of Christians, Jews, and Muslims in the Iberian Peninsula. American Journal of Human Genetics 83: 725 to 736. doi.org/10.1016/j.ajhg.2008.11.007
- Data: Global25 coordinates by Davidski (Eurogenes) and the Moriopoulos Collection. Models, tests and tables: ExploreYourDNA.