By roughly ten generations back, you can have genealogical ancestors from whom you inherited zero autosomal DNA, making you genetically unrelated to them in any measurable sense even though they appear on your family tree. That threshold surprises most people, who assume the connection just gets thinner and thinner but never vanishes. In reality, the randomness of DNA inheritance means the link doesn’t fade gradually like a dimmer switch; it flickers out entirely for specific ancestral lines, and it does so far sooner than our intuitions suggest. The answer also depends on what kind of DNA you’re talking about and what you mean by “related.”
Why DNA Sharing Drops So Quickly
Every time a parent passes DNA to a child, the child gets roughly half of that parent’s genome. Go back one more generation and the expected share from each grandparent is about a quarter. Each additional generation cuts the expected contribution in half again. By the time you reach a tenth-generation ancestor, the expected genetic contribution from any single individual on that level of your tree is vanishingly small: roughly one part in a thousand.
But “expected” is doing heavy lifting in that sentence. The actual amount you inherit from any one ancestor isn’t a clean 50-50 split repeated perfectly at every generation. During the formation of eggs and sperm, chromosomes swap segments with their partner in a process called recombination. Only a fraction of the hundreds of DNA breaks that occur during this shuffling result in actual segment exchanges between chromosomes; most are repaired without swapping material at all.1bioRxiv. A genome-wide atlas of meiotic recombination intermediates reveals distinct modes of DNA repair that direct crossovers away from transcriptionally marked genes The upshot is that the chunks of ancestral DNA you carry get broken into smaller and smaller pieces with each generation, but the breaking isn’t uniform. Some ancestors contribute slightly more than expected; others contribute less. And eventually, some contribute nothing at all.
This randomness is the reason two siblings who share the same parents can get noticeably different results from a DNA test. One sibling might carry a segment from a particular great-great-grandparent while the other lost that segment entirely through the luck of recombination. Scale that variability across ten or more generations and it becomes inevitable that whole ancestral lines drop off your genome.
How Relatedness Fades at Each Step
Geneticists measure shared ancestry by looking for stretches of DNA that two people inherited from a common ancestor, called identical-by-descent (IBD) segments. A recent large-scale study quantified how the probability of sharing any IBD segments at all changes with each degree of relatedness. The results paint a clear picture of the drop-off:2PubMed Central. The rate of identical-by-descent segment sharing between close and distant relatives
- Third cousins: about 98.5% share at least one IBD segment.
- Fifth cousins: only about a third (32.7%) still share detectable DNA.
- Eighth cousins: fewer than 1 in 100 pairs (0.96%) share any IBD at all.
At finer resolution, the data show that sixth-degree relatives (like second cousins once removed) already have a small but real chance of sharing zero segments, and by tenth-degree relatedness the odds are basically a coin flip: around half of pairs share some DNA, and the other half share none.3PubMed Central. The rate of identical-by-descent segment sharing between close and distant relatives Beyond that, the proportion with detectable sharing continues to plummet.
So if you’re looking for a single number, ten generations (equivalent to roughly eighth or ninth cousins) is the neighborhood where genetic relatedness effectively vanishes for most pairs. But it’s not a cliff edge; it’s a rapid slope that starts getting steep around the fifth-cousin mark and is nearly flat by the time you reach tenth cousins.
Genealogical Relatives Versus Genetic Relatives
This is where the question gets genuinely interesting, because “related” can mean two very different things. Your family tree keeps branching backward regardless of whether DNA comes along for the ride. At ten generations back you have up to 1,024 slots for ancestors on your pedigree chart (though many of those slots are filled by the same person due to intermarriage within communities). Every one of those people is your genealogical ancestor. But as we’ve seen, you may carry DNA from only a fraction of them. By around ten generations in the past, you typically have genealogical ancestors from whom you inherited no autosomal DNA whatsoever.4Europe PMC. What is ancestry?
This means two people can be verified genealogical cousins through careful records, church documents, or historical archives, yet share absolutely no detectable DNA. They are related by pedigree but not by genome. For most practical purposes today, when people ask “are we related?” they mean the genetic version. Consumer DNA tests, forensic investigations, and medical genetics all rely on detecting shared DNA segments. If those segments aren’t there, the tools come up empty.
The reverse can also happen. Two people with no known family connection might share a small IBD segment inherited from an ancestor so far back that no paper trail exists. In populations with limited geographic mobility over centuries, this kind of background relatedness is common and can complicate efforts to distinguish true close relatives from people who just happen to belong to the same historically isolated community.
When Theoretical Expectations Don’t Match Reality
The textbook expectation is tidy: each generation halves the shared DNA. In practice, observed genetic sharing between relatives is often slightly higher than the simple halving rule predicts. Research on livestock pedigrees, where parentage records are meticulously kept and can be cross-checked against dense genetic data, found that average genomic relationship coefficients consistently ran about 7 to 9 percent higher than what pedigree-based theory predicted.5PubMed Central. Forensic use of the genomic relationship matrix to validate and discover livestock pedigrees The pattern holds across different relationship categories.
Why the overshoot? Part of it is that real populations are not infinitely large and randomly mating. In any finite group, individuals share more background ancestry than a simple model assumes. Distant cousins marry each other without knowing it, and over time, this quiet overlap inflates the amount of shared DNA slightly above the clean theoretical prediction. In historically small or geographically isolated populations, the effect is larger. In huge, well-mixed populations, it’s smaller but still present. Studies using runs of homozygosity, stretches of the genome where both copies of a chromosome are identical because of shared ancestry, have confirmed that even in large populations with effective population sizes around 10,000, measurable variation in background relatedness persists.6Genetics / PubMed Central. Quantification of inbreeding due to distant ancestors and its detection using dense single nucleotide polymorphism data
The practical takeaway is that the “ten generations” threshold for losing all shared DNA is an average, not a hard boundary. In populations with a history of geographic isolation or cultural endogamy, genetic connections persist longer because ancestors appear multiple times in different branches of your tree, reinforcing the DNA signal. In highly mobile, outbred populations, the connection may fade even faster.
Mitochondrial DNA and the Y Chromosome Play by Different Rules
Everything discussed so far applies to autosomal DNA, the 22 pairs of non-sex chromosomes that get shuffled and halved each generation. But two small pieces of your genome don’t follow those rules at all.
Mitochondrial DNA (mtDNA) passes exclusively from mother to child. Sperm cells carry mitochondria, but those mitochondria lack intact mtDNA and are missing the key protein needed to maintain it, so they contribute nothing to the next generation’s mitochondrial genome.7PubMed Central. Molecular basis for maternal inheritance of human mitochondrial DNA Mitochondrial DNA also doesn’t recombine the way autosomal chromosomes do.8PubMed Central. The bottleneck for maternal transmission of mtDNA is linked to purifying selection by autophagy This means your mtDNA is essentially a copy of your mother’s, which was a copy of her mother’s, stretching back in an unbroken maternal line for thousands of generations. Two people who share a maternal-line ancestor from 500 years ago will still carry nearly identical mtDNA, long after their autosomal connection has vanished.
The Y chromosome works similarly for paternal lines. It passes from father to son with relatively little recombination across most of its length. Y-chromosome studies have traced paternal lineages back hundreds of thousands of years; one African American paternal lineage, designated A00, yielded an estimated common ancestor around 338,000 years ago.9PubMed Central. An African American paternal lineage adds an extremely ancient root to the human Y chromosome phylogenetic tree That’s a connection spanning thousands upon thousands of generations, far beyond anything autosomal DNA could preserve.
So the answer to the title question shifts dramatically depending on which part of the genome you’re asking about. Autosomal relatedness can disappear in as few as ten generations. Mitochondrial and Y-chromosome relatedness can persist for tens of thousands of generations, though those lineages trace only a single thread through your vast ancestry and can’t tell you much about most of the people on your family tree.
Everyone Alive Is Related to Everyone Else
Here is the part that seems paradoxical but is mathematically well-established. The most recent genealogical ancestor common to every living human probably lived only a few thousand years ago. Modeling studies that incorporate realistic population structure, geographic barriers, and migration patterns consistently produce this result.10PubMed. Modelling the recent common ancestry of all living humans Go back a bit further, and you reach a point where every person alive at that time is either an ancestor of everyone alive today or an ancestor of no one alive today. The genealogical connections are that dense.
This doesn’t contradict the ten-generation threshold for DNA sharing. It reinforces it. You have so many genealogical ancestors at deep time depths that the overlap between any two people’s family trees becomes enormous, but the DNA from any specific shared ancestor has long since been diluted to nothing. Two random people alive today share thousands of common ancestors from a few thousand years ago, yet they may share zero detectable autosomal DNA from any of them. They are genealogically related but genetically strangers.
The distinction matters for how you think about heritage. People sometimes assume that having, say, a medieval ancestor in common with someone means they share that ancestor’s DNA. In almost all cases, they don’t. The genealogical connection is real in the pedigree sense but invisible to any genetic test.
Neanderthal DNA and Very Ancient Relatedness
If autosomal connections vanish so quickly, how is it that modern humans of non-African descent still carry detectable Neanderthal DNA from tens of thousands of years ago? The answer is scale. Neanderthal gene flow into the ancestors of modern Eurasians didn’t come from one ancestor contributing a tiny fraction; it resulted from an extended period of interbreeding between two entire populations, concentrated between roughly 50,500 and 43,500 years ago.11Europe PMC. Neanderthal ancestry through time: Insights from genomes of ancient and present-day humans Thousands of individuals over thousands of years contributed Neanderthal segments, and natural selection then acted on those segments, keeping some and purging others. The surviving fragments persist precisely because they were introduced on such a massive scale that random dilution alone couldn’t erase them all.
Neanderthal ancestry in modern genomes amounts to only about 1 to 2 percent, but it’s distributed across different stretches of the genome in different people. When you pool together the Neanderthal segments found across many living humans, they cover a substantial fraction of the Neanderthal genome. Each individual carries only a tiny slice, but collectively the signal persists. This is a fundamentally different situation from the ten-generation fadeout of a single ancestor’s contribution, where no population-wide reinforcement keeps the segments alive.
What Consumer DNA Tests Can Actually Detect
If you’ve taken a consumer DNA test and been matched with distant cousins, you’ve seen the practical side of this question. These tests work by scanning hundreds of thousands of genetic markers and looking for shared IBD segments between users. Forensic genetic genealogy uses the same basic approach, applying dense marker data to infer relationships and, in criminal cases, to identify suspects through their relatives.12Elsevier / Forensic Science International: Genetics. Investigative genetic genealogy: Current methods, knowledge and practice
The detection limit is important for setting realistic expectations. Given that only about a third of fifth cousins share any DNA at all, and fewer than 1 percent of eighth cousins do, most consumer tests reliably identify relatives out to about third or fourth cousins. Beyond that, matches become sparse and uncertain. A “fourth-to-sixth cousin” match on a testing platform might be accurate, or it might reflect background population-level sharing rather than a specific recent ancestor.
Researchers working with ancient DNA have pushed detection further using specialized methods that can identify IBD segments even in degraded, low-coverage genomes.13Nature Genetics. Accurate detection of identity-by-descent segments in human ancient DNA These tools have been used to map family relationships among people buried at the same archaeological site thousands of years ago. But even with advanced methods, the fundamental constraint remains: if recombination has broken the shared segments below a detectable size, no algorithm can find them.
Epigenetic Inheritance Doesn’t Change the Picture Much
You might wonder whether something beyond DNA sequence itself could carry an ancestor’s influence forward across many generations. Epigenetic marks, chemical modifications that sit on top of DNA and affect which genes are active, can sometimes be passed from parent to offspring. In plants, this kind of inheritance is well documented. In mammals, the picture is much murkier. Most epigenetic marks are wiped clean and reset between generations, and the extent to which any environmental influence can genuinely persist across multiple human generations remains unclear.14Europe PMC / Cell. Transgenerational epigenetic inheritance: myths and mechanisms
Even in the most generous interpretation of the evidence, epigenetic inheritance in humans appears limited to a handful of generations at most, not the deep timescales people sometimes imagine. A grandparent’s famine or stress exposure might leave subtle traces in grandchildren through epigenetic pathways, but those traces don’t accumulate or persist the way DNA sequence does. Epigenetics adds a thin, short-lived layer to inheritance rather than extending the reach of relatedness in any meaningful way.
Why Some Populations Stay Genetically Connected Longer
The ten-generation estimate assumes a large, outbred population. For people with ancestry in historically small or isolated communities, such as island populations, certain religious communities, or ethnic groups that experienced historical bottlenecks, the math changes. When your ancestors repeatedly married within the same group, the same individuals appear over and over again in different branches of your tree. This is called pedigree collapse, and it means you effectively have fewer unique ancestors than the theoretical maximum.
The genetic consequence is that DNA from those repeated ancestors gets reinforced. Instead of being diluted to zero by generation ten, segments from a shared ancestor who appears in six different branches of your tree have six independent chances of surviving the recombination lottery. People in these populations tend to share longer and more numerous IBD segments with each other, and their genetic connections extend further back in time before disappearing. This is exactly why forensic genetic genealogy has had some of its most striking successes in populations with strong endogamy: the DNA signal from shared ancestors lasts longer and is easier to detect.
Conversely, in highly cosmopolitan populations with extensive admixture over centuries, the genetic dilution happens faster because there’s less pedigree collapse reinforcing any particular ancestor’s contribution. Your ten-generation ancestors are more likely to be genuinely distinct individuals, meaning each one’s DNA gets fewer chances to survive.
Paternal Leakage and Rare Exceptions to Maternal Mitochondrial Inheritance
The rule that mtDNA passes only through the maternal line holds in the overwhelming majority of cases, but rare exceptions have been reported. A handful of families have shown evidence of biparental mitochondrial DNA inheritance, where paternal mtDNA appears alongside maternal mtDNA in offspring.15PubMed Central. Biparental Inheritance of Mitochondrial DNA in Humans These cases remain controversial and appear to be genuine anomalies rather than a hidden widespread phenomenon. For practical purposes, including ancestry testing and forensic identification, maternal-only mtDNA inheritance is the working assumption, and it holds up for essentially everyone.
The reason these rare exceptions matter to the broader question is that they illustrate how biological rules about inheritance are statistical generalizations, not absolute laws. The same principle applies to the ten-generation autosomal threshold. It’s a robust central tendency, not a guarantee. Some people will lose all detectable DNA from a given ancestor by generation seven. Others, through the luck of recombination and the quirks of population history, might carry a recognizable segment from a twelfth-generation ancestor. The biology is probabilistic from top to bottom.