Comparing DNA Patterns of Sephardi,

           Ashkenazi & Kurdish jews 

by Abraham D. Lavender PhD

Considering the diverse origins of the Israelites, the large numbers of conversions into and out of the Jewish community, and major disasters which have befallen Jewish communities throughout history, what is the genetic composition of Jewish communities today?

From HaLapid, Spring 2005


In the days before genetic testing, the famous Jewish anthropologist Raphael Patai used cephalic indexes and blood groups to show the similarities between Jews and other groups. He and a number of other scholars discussed the mixed origins of the ancient Hebrews and the major extent to which male Hebrews married women from diverse ethnic groups (1971). Shaye Cohen has shown that in antiquity there was not a strong boundary between Jews and Gentiles. By the second century BCE there was a boundary, but it could be crossed, and “gentiles crossed it and became Jews in a variety of ways, whether by political enfranchisement, religious conversion, veneration of the Jewish God, observance of Jewish rituals, association with Jews, or other means” (1999: 342). For the first time, there was now the notion of conversion to Judaism, and religion overcame ethnicity. In the second century CE, the concept of matrilineal descent was begun. Debate continues over the reasons, but Patai notes that it was only after the Roman Exile, when “violation and impregnation of Jewish women by foreign conquerors, invaders, armies, bands, or marauders” became common that Talmudic law changed to recognize matrilineal rather than patrilineal descent (1971: 61).

The genetic composition of the Jewish people became even more mixed after the major dispersions from Israel. After the fall of Israel in 721 BCE (the dispersal of the ten northern tribes), the fall of Judah in 586 BCE (the Babylonian exile and the beginning of the Egyptian Diaspora), and during and after the fall of the Second Commonwealth in 70 CE, “conversion of individuals and groups to Judaism became, if not frequent at least not exceptional.” There are at least three reported cases of groups converting to Judaism: the Kingdom of Abiabene of Iraq, in the first century CE (Wexler, 1996: 28), the Himyars in Yemen in the third to fifth centuries CE (Ben-Zvi, 1957), and the Khazars of Georgia (Eastern Europe) in the ninth century CE (Brook,1999). Patai and Patai note that in pre-Islamic Arabia, “intermarriage between Jews and pagan Arabs was frequent....With the expansion of Islam, Jews intermarried not only with Arabs but also with members of the nations drawn into the Muslim orbit by the Arab conquests” (1980: 103). As parts of Arabia were conquered by Islamic forces, there also were numerous cases of Jews converting, sometimes voluntarily but often involuntarily, to Islam. Stillman (1991), Patai (1971, 1997), Ben-Zvi (1957),  Goitein (1974), and DeFelice (1985) give numerous examples. Wexler, especially, gives examples of conversions going both ways.

There are also examples of descendants of the lost tribes of Israel claiming Jewish origins, with many of them now returning to Judaism. So far there has been little DNA testing, but regarding the Bene Israel Jews of the Bombay area of India, Parfitt concludes that his DNA research “clearly suggests that the Bene Israel are a very ancient, probably Jewish, group” (2003: 11). He also concludes that he has “likely evidence” that the Black Jews of Cochin are descendants of “an early migration of what were probably Jews from the Near East to India in ancient times” (p. 15). The examples of Jews converting to Christianity in the last thousand years or so are too numerous to mention, and there is no question that millions of people who identify as non-Jewish today have Jewish ancestry.  The Jewish community, while bemoaning the loss of so many people, frequently because of discrimination or oppression, also has not sufficiently recognized the large number of people who have joined Jewish communities. The world Jewish population in the first century of the Common Era (CE) was about four to five million, with 1.5 to two million in Israel. Massive conversions, persecutions, and murders have led to drastic population decreases which also have affected the specific genetic distribution of worldwide Jewry (because some Jewish groups and areas have suffered more than others).

The extent of interaction of Jews and non-Jews has been the topic of a number of genetic studies, with the majority of researchers suggesting little interaction and others suggesting substantial interaction. Part of the controversial differences are because of different loci (points on the genetic chain) being tested, different levels of comparisons being utilized, or small samples which can lead to variable results. Space prohibits discussion of all of the genetic studies on this topic, but for an idea of the overall findings, results of some major studies follow.

Santachiara noted in 1993 that mtDNA (female) studies had already been published, by Batsheva Bonne-Tamir et al (1986) and by Tikochenski et al (1991), but that genetic comparisons for male Ashkenazim and Sephardim (Y-chromosomes) had not been done (p. 56). Tikochinski, using Israeli samples, had analyzed twenty-one Ashkenazi women from Eastern Europe and thirty-eight Sephardi women (mostly from Morocco). Her data implied that these Jewish women descended from a diversity of maternal lineages that had been distinct for four to five thousand years. Thomas et al, in 2002, published data on Jewish women in nine geographically separated areas, and concluded that, contrary to non-Jews, there was greater differentiation for mtDNA than for the Y-chromosome, that “cultural practice–in this case, female-defined ethnicity–has had a pronounced effect on patterns of genetic variation” (p. 1417).  By the early 1990s, however, methodological advances were beginning to make it possible for Y-chromosome studies also to be conducted. In 1991, Livshits et al compared twelve pairs of Jewish and non-Jewish populations from the Middle East, North Africa, and Europe with each Jewish/non-Jewish pair sharing the same (or close) geographic area: Yemen, Iran, Iraq; Morocco and Libya; Poland, Russia, and Georgia; Germany and Czechoslovakia; Bulgaria and Turkey/Spain (Turkish Jews were compared to Spanish non-Jews because most Turkish Jews were exiled from Spain by the Inquisition). Their conclusion: modern Jewish populations in general derived from an earlier common gene pool which had undergone relatively little admixture with non-Jewish neighbors after dispersal from Israel. Somewhat differently, Kurdish Jews had experienced considerable interaction with non-Jewish Kurds, and Yemenite Jews may have had a substantial component of different genes from conversion into Judaism (p. 145).

In 1993, Santachiara et al compared eighty-three Sephardim (mostly from Tunisia and Morocco), eighty-three Ashkenazim (mostly from Russia and Poland), and 105 non-Jews from Czechoslovakia, and made comparisons to non-Jews from Lebanon. They found strong genetic affinities between Sephardim and Ashkenazim, almost no relationship with non-Jews from Czechoslovakia, and a very close relationship between Sephardim and Lebanese non-Jews. They found about 23.4% to 28.6% non-Jewish Y-chromosomes in the Ashkenazim, and concluded that this represented about one percent or less of admixture per generation for the centuries the Ashkenazim had lived in Central or Northeastern Europe (p. 63).

In 2000, Hammer et al compared seven Jewish groups (Ashkenazim, Roman, North African, Kurdish, Near Eastern [Iran and Iraq], Yemenite, and Ethiopian) with sixteen non-Jewish groups from similar geographical locations. They concluded that most Jewish groups were similar to each other, and had experienced little genetic admixture with non-Jewish groups. They found a strong genetic similarity between most Jews and Middle Eastern non-Jews. The Palestinians and Syrian non-Jews were most closely related, but Saudi Arabians, Lebanese, and Druze also were close. The authors attributed the genetic closeness to ancient common Middle Eastern origins. They estimated the admixture rate of Ashkenazim (for all haplotypes) to be 22.7% plus or minus 7.8% over a period of about eighty generations (p. 6773). Interestingly, Ethiopian Jews and the Lemba did not match closely with the cluster of Jewish groups (p. 6774).

In 2001, Nebel et al compared three Jewish and three non-Jewish groups from the Middle East: Ashkenazim, Sephardim, and Kurdish Jews from Israel; Muslim Arabs from Israel and the Palestinian Authority Area; Bedouin from the Negev; and Muslim Kurds. They concluded that Sephardim and Kurdish Jews were genetically indistinguishable, but that both were slightly significantly different from Ashkenazim (who were most closely related to the Muslim Kurds). Nebel et al had earlier (2000) found a large genetic relationship between Jews and Palestinians, but in this study found an even higher relationship of Jews with Iraquis and Kurds. They conclude that the common genetic background shared by Jews and other Middle Eastern groups predates the division of Middle Easterners into different ethnic groups (p. 1106).

Interestingly, Nebel et al (2001) also found that the Cohen Modal Haplotype (CMH), considered the most definitive Jewish haplotype, was found among 10.1% of Kurdish Jews, 7.6% of Ashkenazim, 6.4% of Sephardim, 2.1% of Palestianian Arabs, and 1.1% of Muslim Kurds. The CMH and the most frequent Muslim Kurdish haplotype (MKH) were the same on five markers (out of six) and very close on the other marker. The MKH was shared by 9.5% of Muslim Kurds, 2.6% of Sephardim, 2.0% of Kurdish Jews, 1.4% of Palestinian Arabs, and 1.3% of Ashkenazim. The general conclusion is that these similarities result mostly from the sharing of ancient genetic patterns, and not from more recent admixture between the groups (p. 1099). Rabbi Yaakov Kleiman has suggested that the CMH is “likely the marker of the Jews’ and Arabs’ shared Patriarch, Abraham” (2004: 20), but much more analysis is needed on the CMH in populations throughout the world.

In 2004, Behar et al compared data from Ashkenazi groups in ten different European areas (France, Germany, the Netherlands; Austria-Hungary, Byelorussia, Lithuania, Poland, Romania, Russia, and Ukraine ) with data from non-Jewish groups in seven different countries (France, Germany, Austria, Hungary, Poland, Romania, and Russia). They found that nine of the Jewish groups were similar, with low rates of admixture with non-Jewish groups, but that these Ashkenazi groups were closely related to non-Ashkenazi Jews and to some non-Jewish Near Eastern groups. Within Europe, these authors suggested an admixture rate of 5-8% for all the Jewish nationality groups except Dutch Jews who had an admixture rate of 46.0% plus or minus 18.3%. This supposedly resulted from a long history of relative tolerance from non-Jews, with Jewish women marrying non-Jewish men (p. 362).

Researchers frequently have used haplogroups to measure the genetic distance between various groups in the world, and combined data allows the comparison of these three Jewish groups–Ashkenazim, Sephardim, and Kurdish Jews–with sixty-nine non-Jewish Eurasian populations. The lower the number, the less genetic distance there is between the groups, i.e., the closer the groups are related genetically. The lowest number for the Jewish groups (closet relationship) was 18, and the highest number (least relationship) was 88. It is important to note that these relative average genetic distances will differ as more samples are added, as data from different methodological techniques are evaluated, and as different judgments are made on adjusting different studies to make them comparable to each other. Genetic research is a relatively recent, and rapidly evolving, area of research, and there will be constant refinements and adjustments as more research is added.

Of all the groups, the Ashkenazim are most closely related, in order, to Palestinian Arabs (18), Muslim Kurds (21), Cypriots (22), Greeks (23), Kurdish Jews (25), Bedouin (26), Sephardi Jews (27), Egyptians (27), Turks (28), and Pakistani Parsi (31). Sephardim are most closely related to Italians (18), Turks (20), Ossetians, Georgia (20), Kurdish Jews (22), Muslim Kurds (24), Greeks (24), Armenians (26), Cypriots (26), Ashkenazi Jews (27), and Pakistani Parsi (28). Kurdish Jews are most closely related to Sephardi Jews (22) and Muslim Kurds (22), Pakistani Parsi (23), Ashkenazi Jews (25), Turks (26), Palestinian Arabs (28), Ossetians (30), Cypriots (31), Greeks (32), and Armenians (35). On the other hand, for nineteen Central and Eastern European populations, the Ashkenazim averaged 55.1, the Sephardim averaged 48.2, and the Kurdish Jews averaged 55.5. The Georgia Ossetians (X=29.7) and the Romanians (X=33.0) were the closest to all three of the Jewish groups. For thirteen western European populations, the Ashkenazim averaged 71.6, the Sephardim averaged 47.9, and the Kurdish Jews averaged 63.0. The Sephardim were most closely related to the Southern Portuguese (33), Dutch (36) and French (36), and Northern Portuguese (40). None of these were particularly close, but they were much closer than they were for Ashkenazim or Kurdish Jews.

The genetic distances shown in the previous paragraph are in general agreement with the studies reported in this article, showing that all three Jewish groups are relatively closely related to each other. The close relationship between Sephardim and Kurdish Jews is possibly at least partly a result of the significant interaction between the Jews of Iraq and the Jews of Spain and North Africa, especially from the eighth to the tenth centuries CE. The three Jewish groups differ in their rankings with their closest ten groups, but generally the differences in rankings for the closet ten groups are small and subject to changes in ranking as more samples are added. All three Jewish groups are closely related to Kurdish Muslims, the closest neighbors of the Kurdish Jews. Kurdish Jews were close to Muslim Kurds, but so were Ashkenazim and Sephardim, suggesting that much if not most of the genetic similarity between Jewish and Muslim Kurds is from ancient times. Considering their physical closeness, however, it is reasonable to believe that there has been some genetic admixture not picked up because the two groups started with similar genetic patterns.

Ashkenazim are not closely related to their Central and Eastern European neighbors or to any group outside the Middle East or Near East. Sephardim are more closely related to their neighbors than are the Ashkenazim, but the Sephardim still are much more closely related to the other two Jewish groups, the other Middle Easterners, and the Mediterraneans than they are to their western European neighbors. The Jewish community in the Netherlands is the most obvious example of genetic admixture, a pattern which will be seen more often due to major increases in intermarriages. The other side of the coin, the extent to which Jewish genetic patterns have entered non-Jewish groups, is also a topic which needs much more specific research.        

Recently, there also has been specific archaeological interest given to the origins of the early Israelites, a controversial topic within archaeology. Dever has recently written that the proto-Israelites consisted of local pastoral nomads, refugees fleeing Egyptian injustice, social rebels and social “bandits,” urban dropouts escaping exploitation, corruption, and inefficiency, former soldiers and mercenaries, and others (pp. 181-182). He concludes that the solidarity formed by these people was ideological. not biological.




Behar, Doron et al. “Contrasting Patterns of Y Chromosome Variation in Ashkenazi Jewish and Host Non-Jewish European Populations.” Human Genetics, Volume 114, 2004, pp. 354-365.

Ben-Zvi, Itzvak. The Exiled and the Redeemed. Philadelphia: The Jewish Publication Society, 1957.

Bonne-Tamir, B., et al. “Human Mitochondrial DNA Types in Two Israeli Populations–A Comparative Study at the DNA Level.” American Journal of Human Genetics, Volume 38, 1986, pp. 341-351.

Brook, Kevin Alan. The Jews of Khazaria. Northvale, N.J.: Jason Aronson, Inc., 1999.

Cohen, Shaye J.D. The Beginnings of Jewishness: Boundaries, Varieties, Uncertainties. Berkeley: University of California Press, 1999.

DeFelice, Renzo. Jews in an Arab Land: Libya, 1835-1970. Austin: University of Texas Press, 1985.

Dever, William G. Who Were the Early Israelites and Where Did They Come From? Grand Rapid, Mich.: William B. Eerdmans Publishing Company, 2003.

Hammer, Michael F. et al. “Jewish and Middle Eastern non-Jewish Populations Share a Common Pool of Y-Chromosome Biallelic Haplotypes.” PNAS, Volume 97, Number 12, June 6, 2000, pp. 6769-6774.

Kleiman, Rabbi Yaakov. DNA & Tradition: The Genetic Link to the Ancient Hebrews. Israel: Devora Publishing, 2004.

Livshits, Gregory, Robert R. Sokal, and Eugene Kobyliansky. “Genetic Affinities of Jewish Populations.” American Journal of Human Genetics, Volume 49, 1991, pp. 131-146.

Nebel, Almut, et al. “High-Resolution Y Chromosome Haplotypes of Israeli and Palestinian Arabs Reveal Geographic Substructure and Substantial Overlap With Haplotypes of Jews.” Human Genetics, Volume 107, 2000, pp. 630-641.

Nebel, Almut, et al. “The Y Chromosome Pool of Jews as Part of the Genetic Landscape of the Middle East. “American Journal of Human Genetics, Volume 69, 2001, pp. 1095-1112.

Parfitt, Tudor. “Genetics and Jewish History in India: The Bene Israel and the Black Jews of Cochin.” The Journal of Indo-Judaic Studies, Number 6, 2003, pp. 7-18.

Patai, Raphael. Jadid Al-Islam: The Jewish “New Moslems” of Meshhed. Detroit: Wayne State University, 1997.

Patai, Raphael. Tents of Jacob: The Diaspora–Yesterday and Today. Englewood Cliffs, N.J., Prentice Hall, 1971.

Patai, Raphael, and Jennifer Patai. The Myth of the Jewish Race. Detroit: Wayne State University, 1989.

Santachiara, A.S., et al. “The Common, Near-Eastern Origin of Ashkenazi and Sephardi Jews Supported by Y-Chromosome Similarity.” Annals of Human Genetics, Volume 57, 1993, pp. 55-64.

Stillman, Norman. The Jews of Arab Lands in Modern Times. Philadelphia: The Jewish Publication Society, 1991.

Thomas, Mark G., et al. “Founding Mothers of Jewish Communities: Geographically Separated Jewish Groups Were Independently Founded by Very Few Female Ancestors.” American Journal of Human Genetics, Volume 70, 2002, pp. 1411-1420.

Tikochinski, Y., et al. “mtDNA Polymorphism in Two Communities of Jews.” American Journal of Human Genetics, Volume 48, 1991, pp. 129-136.


ABRAHAM D. LAVENDER, President of SCJS, is Professor in the Department of Sociology and Anthropology, Florida International University, Miami.  He has addressed Society conferences, as well as written articles and papers, on related topics.

















Society For Crypto Judaic Studies