A 7% shared-DNA match most likely traces back to a common ancestor about three to five generations ago, with four generations (your great-great-grandparents’ level) being the statistical sweet spot. That said, DNA inheritance is not perfectly predictable, and the same 7% figure can represent several different relationship types. Understanding why the number is fuzzy, and what it can and cannot tell you about your family tree, is worth a closer look.
The Halving Rule and Where 7% Fits
Each generation you move away from an ancestor, you share roughly half as much DNA with them as the generation before you did. You get about 50% of your DNA from each parent, about 25% from each grandparent, about 12.5% from each great-grandparent, and about 6.25% from each great-great-grandparent. That 6.25% figure is the theoretical expectation for someone four generations removed from a shared ancestor, making it the closest textbook match to a 7% result.
But “theoretical” is doing a lot of work in that sentence. DNA is not passed down in perfectly equal slices. Each time a parent passes DNA to a child, the process involves a random reshuffling of genetic material. One grandchild might inherit 27% of a grandparent’s DNA while a sibling inherits 22%. By the time you get to great-great-grandparents, the randomness compounds. Real-world percentages for four-generation relationships can land anywhere from roughly 3% to 12%, depending on luck. So a 7% match is consistent with four generations back, but it does not rule out three or five.
Which Relationships Match Around 7%
A shared-DNA percentage doesn’t point to one specific relationship. It points to a cloud of possibilities. At around 7%, the main candidates include:
- First cousins once removed: You and the child (or parent) of your first cousin. The expected share is 6.25%, but observed values typically range from about 3.3% to 8.5%.
- Half first cousins: You share one grandparent instead of two. Expected share is also 6.25%, with a similar spread.
- Second cousins (high end): You share a set of great-grandparents. The expected share is about 3.1%, but a lucky roll of genetic recombination can push this up toward 7% in some pairs.
- First cousins (low end): The expected share for full first cousins is 12.5%, but some pairs share as little as 7% to 8%. This is unusual but not unheard of.
- Great-great-grandparent to great-great-grandchild: A direct-line relationship four generations apart. Expected share is 6.25%.
The overlap between these categories is the central frustration of genetic genealogy at this percentage range. You cannot look at “7%” alone and know whether the match is a half first cousin, a generous second cousin, or a stingy first cousin. The DNA percentage narrows the field; the family tree research closes the gap.
Why the Same Percentage Can Mean Different Generations
The randomness of DNA inheritance is not just noise. It has a biological cause. When egg and sperm cells are formed, chromosomes swap segments with their partner chromosome in a process called recombination. Which segments get swapped is essentially random. This means two siblings who share the same parents will each pass down different chunks of an ancestor’s genome to their own children. Over a few generations, the variability accumulates to the point where two people who are technically the same degree of relatedness can share quite different amounts of DNA.
Consider two pairs of second cousins. One pair might share 4.5% of their DNA, the other might share 1.8%. Both are equally “second cousins” in a genealogical sense, but genetically they look like they could be different relationship types. This is why testing companies report a range of possible relationships rather than a single answer. At 7%, you are sitting in a zone where multiple relationship categories overlap, and no amount of staring at the number will resolve the ambiguity without tree-building.
Counting Generations to a Common Ancestor
When people ask “how many generations back,” they usually mean: how many steps from me to our shared ancestor? The generation count depends on which relationship you’re actually in. For first cousins once removed, one person is two generations from the common ancestor and the other is three. For second cousins, both are three generations from the shared ancestor. For half first cousins, both are two generations from the one shared grandparent.
So a 7% match could mean the common ancestor is as close as two generations back (a shared grandparent, if the match is a half first cousin) or as far as five generations in rare cases. The most probable range is three to four generations. If you’re trying to pin it down, the DNA alone gives you odds, not certainty. The most productive next step is comparing family trees with the person you’ve matched. Even partial trees with locations and surnames can quickly eliminate whole relationship categories.
What Centimorgans Tell You That Percentages Don’t
Most DNA testing platforms report shared DNA in centimorgans (cM) as well as a percentage. The total human genome measures roughly 6,800 cM across all 22 pairs of autosomal chromosomes, so 7% works out to about 476 cM. Centimorgans are useful because genealogy reference charts, shared-DNA tools, and community databases all use them as their standard unit. If you plug 476 cM into a tool like the Shared cM Project (a large crowd-sourced dataset of known relationships and their DNA overlap), you get a probability distribution across relationship types rather than a single answer.
At around 476 cM, those probability distributions confirm what the percentage suggests: first cousin once removed, half first cousin, and great-aunt or great-uncle are the most likely categories, with second cousin and first cousin sitting at the edges. The value of the centimorgan figure is that it lets you cross-reference against large datasets where thousands of people with confirmed relationships have reported their actual shared DNA. That empirical data is far more useful than theoretical percentages alone, because it accounts for the real-world scatter caused by recombination.
How Endogamy Inflates the Number
Endogamy refers to the long-standing practice of marrying within a defined community, whether for geographic, religious, cultural, or ethnic reasons. Populations with a history of endogamy include Ashkenazi Jewish communities, French Canadians, certain Polynesian islands, Amish and Mennonite groups, and many others. In endogamous populations, people are related to each other through multiple ancestral lines, not just one. You and your match might be third cousins through one branch and fourth cousins through another, and the DNA from both connections adds up.
The practical result is that endogamy inflates shared-DNA percentages. A match that looks like 7% in a non-endogamous context might actually represent a more distant relationship, say five or six generations back, because the shared DNA is the sum of multiple smaller connections rather than one close one. If you know your family comes from an endogamous population, you should generally assume that your DNA matches are further back than the standard charts suggest. Some genetic genealogists recommend mentally shifting the relationship estimate one or two categories more distant when working within endogamous lines.
This is one of the more common sources of confusion for people using consumer DNA tests. A 7% match in an endogamous population might correspond to what would normally look like a 3% or 4% match in an outbred population, once you account for the multiple overlapping lines of descent. The shared DNA is real, but it is stacked up from several relationships rather than concentrated in one.
Half Relationships and How They Shift the Math
Half relationships arise when two people share one common ancestor in a pair rather than two. Half siblings share one parent instead of two. Half first cousins share one grandparent instead of a pair. These relationships share roughly half the expected DNA of their full equivalents, which means they often land in the same percentage range as a more distant full relationship.
At 7%, a half first cousin (sharing one grandparent, expected 6.25%) looks almost identical to a first cousin once removed (sharing a pair of grandparents, expected 6.25%). The DNA cannot distinguish between these on its own. What differs is the family structure: a half first cousin involves a half-sibling relationship somewhere in the tree, while a first cousin once removed involves a generational offset. If you’re trying to figure out which one applies, look at the age difference between you and your match. A generational offset usually shows up as a significant age gap, while half cousins tend to be closer in age. This is a rough guide, not a rule, but it helps.
Unknown half relationships are more common than many people expect. Discovering that a match is a half sibling or half first cousin rather than a more distant full relative is one of the more emotionally charged outcomes of consumer DNA testing. If your 7% match does not fit neatly into any known branch of your family tree, the possibility of a half relationship is worth considering before you assume the connection is several generations back.
When the Percentage Comes from Multiple Segments Versus One Large One
Shared DNA is distributed across your chromosomes in segments. A close relative tends to share a few large, continuous segments, while a more distant relative tends to share several smaller ones. The total may add up to the same percentage, but the segment pattern carries information about how far back the relationship is.
If your 7% match is concentrated in two or three large segments, that is more consistent with a closer relationship, perhaps a first cousin who happened to share less DNA than average, or a half great-aunt. If the same 7% is spread across eight or ten smaller segments, that pattern points toward a more distant connection where the original ancestral DNA has been broken up by several rounds of recombination. Most testing platforms will show you a chromosome browser or at least report the number and size of shared segments, which gives you a second axis of information beyond the raw percentage.
Segment size also matters because very small segments (below about 7 cM) are often “false matches,” meaning they look like shared DNA but are actually just common patterns in the population rather than evidence of a recent shared ancestor. Reputable testing companies filter out most of these, but when you compare results across platforms, keep in mind that different companies use different thresholds for what counts as a real shared segment.
Practical Steps After Getting a 7% Match
If you have received a 7% DNA match and want to figure out who this person is in your family, the most productive path combines several approaches. Start by looking at the relationship predictions your testing platform provides. These are statistical best guesses, and they are usually right about the broad category even if the specific label is uncertain. A 7% match will typically show up as something in the first-to-second-cousin range.
Compare your shared matches. Most platforms let you see which other people both you and your match are related to. Clusters of shared matches can help you figure out which branch of your family tree the connection runs through. If you and the mystery match both share DNA with your known maternal second cousins but none of your paternal ones, that narrows the search to your mother’s side of the tree.
Build your tree as far back as you can, ideally to the great-great-grandparent level. At 7%, the common ancestor is most likely at or near that generation. If your match has also built a tree, look for overlapping surnames, geographic regions, or time periods. Even one shared surname in the right place can crack the case open. If neither of you has a deep tree, vital records, census data, and church records from the relevant era are the standard genealogical tools for extending your lines back.
Be prepared for the possibility that the answer is not straightforward. The match could involve a non-paternity event, an informal adoption, a half relationship, or an ancestor who used a different name. DNA doesn’t lie about the biology, but it doesn’t know anything about the family stories you grew up with. A 7% match that doesn’t fit your known tree is not a testing error. It is information about your actual biological family, which may or may not align perfectly with your documented one.
Why Consumer Tests Sometimes Show Different Percentages for the Same Match
If you and a relative both test with multiple companies, you may notice that your shared-DNA percentage differs slightly between platforms. This happens because each company uses its own reference panels, algorithms, and segment-length thresholds to calculate shared DNA. One platform might report 7.1% while another reports 6.4% for the same pair of people. The underlying biology has not changed; the measurement tools are just calibrated differently.
These differences are usually small enough that they do not change the relationship category. A match that shows 7% on one platform and 6% on another is still in the same generational range. Where it can matter is at the boundary between categories. If one platform reports 7% and another reports 5%, the first might suggest “first cousin once removed” while the second leans toward “second cousin.” Neither is wrong exactly; they are using different rulers to measure the same thing. When in doubt, trust the platform where both you and the match tested on the same chip, since comparing results across platforms introduces extra noise.
For the question of how many generations back a 7% match falls, the cross-platform variation reinforces the broader point: treat the percentage as an estimate within a range, not a surgical measurement. The most accurate genealogical conclusions come from combining the DNA data with documentary evidence, shared-match analysis, and a willingness to consider relationship types you might not have initially expected.