They dress plainly, marry within the fold, and have kept to themselves for almost five centuries. The Amish, the Mennonites and the Hutterites all descend from a single religious movement born in the heat of the Reformation. Yet when you look at their DNA, they do not form one people. They form at least two, and the split runs along a line drawn by history rather than by faith.
One movement, several peoples
The story begins in Zurich in 1525, when a group of radical reformers rejected infant baptism and insisted that baptism should be a choice made by adults. Their opponents called them Anabaptists, the re-baptizers. Pacifist, communal in spirit and unwilling to bow to either Catholic or Protestant authorities, they were persecuted across Europe and pushed steadily outward from their homelands.
From that common root, three main branches grew. The Swiss Brethren spread through the Alpine cantons and the upper Rhine. In 1693 a stricter faction led by Jakob Ammann broke away to become the Amish, while the rest remained Swiss Mennonites. A second branch formed in the Low Countries and northern Germany around the teaching of Menno Simons, the man who gave the Mennonites their name. A third, the Hutterites, organised around Jakob Hutter and adopted full community of goods.
What matters for genetics is where each branch went next. The Swiss and south German Anabaptists who became the Amish and the Swiss Mennonites moved relatively directly to Pennsylvania during the 1700s. The Dutch and north German Mennonites took a longer road. They settled first in the Vistula delta of Prussia, then accepted Catherine the Great's invitation to farm the newly conquered lands of southern Russia and Ukraine from 1789 onward. Only in the 1870s did large numbers of these so-called Russian Mennonites move on again, to Manitoba, Kansas and later to Mexico, Paraguay, Bolivia, Argentina and Brazil. The Hutterites followed a similar eastern path through Russia before reaching South Dakota and Manitoba.
Two routes, two histories. As we will see, the DNA still remembers both.
The founder effect is the real story
Before comparing the branches, it helps to understand what makes these groups genetically distinctive in the first place. The answer is not some exotic admixture. It is the founder effect, the genetic fingerprint of a population built from a small number of ancestors who then married mostly among themselves for generations.
The numbers are striking. The South Dakota Hutterites in one well known research cohort trace back through a thirteen generation pedigree to only sixty-four founders. When researchers sequenced the exomes of 7221 Old Order Amish from Lancaster County and compared them with an equal number of British participants from the UK Biobank, the Amish carried only about fourteen percent as many distinct variants. That collapse in diversity has medical consequences: roughly one Amish person in twenty carried an enriched pathogenic variant from the screening panels used. Researchers have even measured a reduced rate of new mutations in the Amish, a sign of how quickly such isolated histories reshape a genome.
The mechanism behind all of this is genetic drift acting on a tiny founding pool, amplified by endogamy and long runs of homozygosity. It does not add new ancestry. It reshuffles and concentrates what the founders happened to carry. Keep that idea in mind, because it explains a trap we will meet at the end.
Reading the northern branch in G25
To place these populations on the European map we can use Global25, the coordinate system published by Davidski of the Eurogenes blog, which represents each sample as a point in a twenty-five dimensional space derived from genome-wide data. A modelled profile for a northern, Russian-Mennonite type sample sits very close to the present day populations of the North Sea zone.
Measured as scaled Euclidean distance, the modelled northern profile lands at 0.0099 from the Dutch average, 0.0101 from the Danish, and 0.0117 from Hamburg Germans. A simple mixture model returns a blend dominated by Danish and German references with a small Scandinavian pull, and a fit residual of only 0.0031, which is a tight match. There is no meaningful non-European component at all. A separate K36 admixture breakdown tells the same story from another angle: the profile is led by North Sea, North Atlantic and Fennoscandian components, with the Iberian and Italian readings being the usual generic Neolithic-farmer background that every northwest European carries rather than any real southern input.
You can paste the modelled coordinates into a tool such as Vahaduo to reproduce the position yourself.
One detail deserves an honest note. Consumer ancestry reports for people from this branch often show a small Eastern European reading, with Russia and Ukraine flagged, alongside the dominant Dutch and German signal. This is exactly what you would expect as a faint echo of the eighty years the group spent farming in the Russian Empire. We cannot prove from a single percentage whether it reflects a real, modest Slavic contribution or a quirk of the reference panels used by these tests. The careful statement is that it is consistent with the documented sojourn in Russia, not that it is firm evidence of large scale mixing. Genome-wide studies of Argentinean and Brazilian Mennonites found their closest match to be central European reference samples, with raised inbreeding coefficients and no sign of Native American or African admixture.
The southern, Swiss branch
The Amish and the Swiss Mennonites tell a different genetic story. Consumer reports for this branch anchor firmly on the Bernese Alps and the central Swiss plateau, and they carry no Eastern European reading at all. Their ancestry sits in the Swiss, Alsatian and south German space, precisely the homeland the Swiss Brethren never strayed far from before crossing the Atlantic.
We can put a number on the gap. Using real reference populations from the Moriopoulos 2025 modern collection as stand-ins for the southern homeland, the modelled northern profile sits at 0.0385 from Swiss Germans, 0.0327 from Alsatians of the Bas-Rhin, and 0.0320 from south German samples of Baden-Wurttemberg. Compared with the 0.0099 that separates it from the Dutch, the northern branch is several times more distant from the Swiss core than from the North Sea coast. Averaged into two poles, the northern profile sits at 0.0075 from the northern reference centre and 0.0281 from the southern one.
The modelled northern profile sits close to the Dutch, Danish and north German references and several times further from the Swiss and south German ones. The contrast between the poles is the robust signal; the exact split among the closely related northern references is not.
Two honest caveats belong with this figure. First, the northern profile here is a modelled point rather than a value read off genotyped church members, so it illustrates the position rather than proving it. Second, because the Dutch, Danish and German references are themselves close neighbours, the precise division of the mixture among them is not reliable; what holds firmly is the wide gap between the northern and southern poles.
The drift trap, not antiquity
There is a tempting mistake waiting whenever an isolated group is plotted on a principal component analysis. Founder populations often drift away from their neighbours and can appear as outliers, sitting oddly apart from the cloud of related samples. It is easy to read that separation as a sign of deep, unusual or ancient ancestry. For Anabaptist groups it is nothing of the kind.
Their distinctive position comes from the founder effect described earlier: a handful of ancestors, generations of marrying within the community, and the slow statistical wandering of allele frequencies that drift produces. The raw material is ordinary northwest and central European ancestry. Drift simply concentrated a particular sample of it. An outlier on a plot is a measure of isolation and small founding size, not of antiquity or exotic origin.
What the DNA remembers
The genetics of the Amish, the Mennonites and the Hutterites turns out to be a faithful record of their separate journeys. The Swiss Brethren who became the Amish and the Swiss Mennonites carry the signature of the Bernese plateau and the upper Rhine, close to Alsatian and south German populations. The Dutch and north German Mennonites carry a North Sea signature with a faint eastern shading that lines up with their long stay in the Russian Empire. Both branches share the same deeper truth: their most distinctive genetic feature is not where their ancestry came from, which is plainly European, but how few people it came through.
Scientific sources
The figures cited above are drawn from the published literature on Anabaptist founder genetics. The main references are listed below.
- Exonic Variation and Its Clinical Impact in 7221 Old Order Amish. American Journal of Medical Genetics Part A, 2025. doi:10.1002/ajmg.a.64212. The source of the fourteen percent variant figure relative to the UK Biobank and the finding that about 5.2 percent of Amish carry an enriched pathogenic or likely-pathogenic variant.
- Associations of genome-wide and regional autozygosity with 96 complex traits in old order Amish. BMC Genomics, 2023. doi:10.1186/s12864-023-09208-5. Documents the Lancaster Amish as a recent founder population roughly fourteen to fifteen generations removed from its European founders, with elevated runs of homozygosity.
- A population-specific reference panel empowers genetic studies of Anabaptist populations. Scientific Reports, 2017. doi:10.1038/s41598-017-05445-3. The Anabaptist Genome Reference Panel, built from whole-genome sequencing of 265 people of Amish and Mennonite ancestry, showing reduced diversity relative to European reference panels.
- An estimate of the average number of recessive lethal mutations carried by humans. Genetics, 2015. arxiv.org/abs/1407.7518. Describes the South Dakota Hutterite thirteen-generation pedigree of 3657 individuals tracing back to 64 founders.
- De novo mutations across 1,465 diverse genomes reveal mutational insights and reductions in the Amish founder population. PNAS, 2020. doi:10.1073/pnas.1902766117. Reports a reduced rate of new mutations in the Amish founder population.
- Genomic continuity of Argentinean Mennonites. Scientific Reports, 2016. doi:10.1038/srep36392. Finds European ancestry closest to central European reference samples, raised inbreeding coefficients, and no Native American or African admixture.
- Pathogenic Variants in Mennonites From Southern Brazil. 2025. PMC12779222. Attributes the Mennonite genetic profile to several demographic bottlenecks and genetic drift.
Population coordinate data: Global25 scaled coordinates, produced by Davidski of the Eurogenes blog, and the Moriopoulos 2025 Modern Population Collection. The northern Mennonite Global25 point used here is a modelled profile included for illustration rather than a value derived from genotyped community members.
