In the 23andMe DNA Relatives feature, we estimate the genealogical relationship between two individuals. We do this by comparing your autosomal DNA (chromosomes 1-22) and X chromosome(s) with other 23andMe members who are participating in the DNA Relatives feature.
While the autosomal DNA is inherited in the same way for both genders, the X chromosome is not:
- Autosomal DNA is inherited equally from both parents for both men and women.
- The X chromosome is a sex chromosome; women inherit an X chromosome from each parent, while men only receive an X chromosome from his mother. In men, the X chromosome is paired with the Y chromosome - which is only inherited from his father.
The Y chromosome is not used by the 23andMe DNA Relatives feature to detect matches.
We say that two individuals share DNA when both individuals inherited the same DNA from the same ancestor. For example, you and your sister share DNA that you both inherited from the same parent. You and your first cousin share DNA inherited from your mutual grandparents. The 23andMe DNA Relatives feature uses patterns of DNA sharing to estimate relationships.
In this article, we will discuss in more detail:
- What we mean by "relative"
- How the DNA Relatives feature detects a genetic relative
- The probability that the DNA Relatives feature will be able to detect a genetic relationship
- How the DNA Relatives feature predicts a genetic relationship
- The average percent DNA shared by relationship
- Common questions
What we mean by relative?
Strictly speaking, two individuals are relatives if they have a common ancestor. Therefore, all humans are relatives by definition. However, in practice, the word relative is restricted to individuals who share recent ancestors. For example, you and your first cousin have the same grandparents, while you and your second cousin have the same great-grandparents. When we say relative, we are referring to individuals who share recent ancestors, and when we say that two individuals are unrelated in this help article, we mean that their common ancestor is 9 or more generations back.
Finding genetic relatives
The first step towards finding your genetic relatives is searching our database for individuals with whom you share DNA. This is not a trivial task since we all share many small DNA segments with each other. However, on average you share longer stretches of DNA with your relatives than with unrelated individuals.
Our simulations have concluded that we can confidently detect related individuals if they have at least one continuous region of matching SNPs (Single Nucleotide Polymorphisms) that is longer than our minimum threshold of 7cM (centiMorgans) long and at least 700 SNPs.
Using that criteria, we estimate the likelihood of detecting a genetic relationship with a 3rd cousin to be approximately 90%, The table below shows the average likelihood of detecting cousins of different degrees:
|Probability of detecting
|1st Cousin or closer
|6th Cousin and beyond
The more distant the relationship, the less likely it is that a genetic relationship will be detected. Distant relatives are less likely to be detected because fewer of them share DNA segments long enough to meet our threshold. Even though the likelihood of detecting distant cousins is low, you will most likely have many distant cousins listed in the DNA Relatives feature. This is because you have more distant cousins than close ones.
Predicting a Relationship
The DNA Relatives feature estimates a predicted relationship and range using the number of segments and percent DNA shared.
In general, you share more segments and a larger percentage of DNA with close relatives than with distant ones.
A notable exception to this general rule is a parent and child relationship. If you are female, you share 23 segments with each parent, and each segment spans the entire chromosome. If you are male, you share 22 segments with father (since the DNA Relatives feature does not use the Y chromosome) and 23 segments with your mother. While you might share more segments with your brother or first cousin than with your father, those segments will be shorter.
The vast majority of genetic relatives found by the DNA Relatives feature share a common ancestor within the last five to ten generations. A few may be more distantly related. A predicted relationship is provided to help you locate your likely recent common ancestor. The predicted relationship uses the percent DNA shared and the number of segments to predict your likely relationship with that relative. In some cases, multiple relationships overlap in the number of segments and percentage DNA shared. Due to this overlap, some relationships can be difficult to pinpoint or can show a predicted relationship that differs from your known relationship with that relative. For distant relationships, we report uncertainty by also showing a relationship range.
We label individuals whose degree of relationship is difficult to estimate due to the small amount of DNA shared as "distant cousins".
Lastly, if the predicted relationship is not what you know the relationship to be, you can set the known relationship instead. You can do this by clicking on the predicted relationship in the DNA Relatives feature.
Percent DNA Shared by Relationship
To help in interpreting your results in the DNA Relatives feature, the following summarizes the average percent (%) DNA shared for different types of relationships. You may notice that several relationships share the same average percent (%) DNA.
|Average % DNA Shared
|Parent / Child
|Varies by specific relationship
|Grandparent / Grandchild
Aunt / Uncle
Niece / Nephew
|Varies by specific relationship
|7.31% - 13.8%
|1st Cousin once removed
|3.3% - 8.51%
|2.85% - 5.04%
|2nd Cousin once removed
|0.57% - 2.54%
|0.3% - 2.0%
|0.07% - 0.5%
You can find a list of the average percent of autosomal DNA shared by close relatives on the International Society of Genetic Genealogy (ISOGG) website.
 The observed ranges by known relationship reported by 23andMe customers on the International Society of Genetic Genealogy (ISOGG) website.
Can DNA Relatives distinguish between a full and half-sibling?
Yes, your relationship to your sibling would be labeled as "Siblings" if full or "Half-siblings" if partial. The DNA Relatives feature uses the length and number of identical segments to predict the relationship between people. Full siblings share approximately 50% of their DNA, while half-siblings share approximately 25% of their DNA.
Why isn't an expected relative listed in DNA Relatives?
There are a couple of reasons why an expected relative would not be listed in the DNA Relatives feature:
- Your relative is not participating in the DNA Relatives feature. Participation in this feature is optional. If your relative has selected not to participate, he or she will not be listed as a relative. Learn more.
- Your list has not yet been updated to show the relative. There may be a delay between when a genetic relative is added to the DNA Relatives database and when he or she can be seen in your list. Frequently visiting this feature will help keep your list up-to-date.
- A genetic relationship could not be found. This is more likely to happen if the expected relative is a distant cousin with whom you share short segments. There is nothing wrong with sharing short segments with distant cousins. In fact, this is expected because of the randomness associated with how DNA is passed over generations. However, confidently detecting short segments is harder than long ones, and we might not be able to find all of the short segments. The probability of detecting a relationship decreases as the number of generations between two individuals increases. We can detect a first cousin or closer relationship with a probability of approximately 100%. Learn more.
Why is my half-sibling predicted as an aunt/uncle/grandparent?
The DNA Relatives feature estimates a Predicted Relationship and range using the number of segments and percent DNA shared. Different relationships can share the same percentage of DNA.
Half-siblings share the same percentage of DNA as:
- An uncle and a niece or nephew
- An aunt and a niece or nephew
- A grandparent and grandchild
If a predicted relationship doesn't match your known relationship, you can update the relationship by clicking on the match and then Edit (located in the Relationship section). Learn more.
DNA Relatives says we are third cousins, but I know our relationship must be more distant. How can this happen?
For more distant relationships, the Predicted Relationship should be treated as an estimate, and you should use the Relationship Range as the bounds of the relationship. The reason lies in the random nature of DNA transmission across generations. On average, you will share more DNA with a third cousin than fourth; however, a few fourth cousins could share more DNA with you than a particular third cousin shares with you. As an example, a maternal fourth cousin might share more autosomal DNA with you than a paternal third cousin.
Because of DNA’s random nature, pinpointing the exact relationship for more distant cousinships can be difficult. Additionally, some populations have higher than average DNA sharing, which can skew the Predicted Relationship to a closer relationship. Learn more.
Are you categorizing cousins differently for people with Ashkenazi ancestry?
Yes, we are making relationship predictions that take into account the higher than average DNA sharing found in people with Ashkenazi ancestry. In many of these cases, this results in a "distant" cousin prediction.
How can a child have a match that doesn't appear in either of their parents' DNA Relatives?
It does happen that a child sometimes shows a match that's not visible in either parent. In these cases, it's almost always the case that the match is indeed present in one of the parents, but the segment identification algorithm has missed the segment in the parent. This typically happens with short segments, where it is more likely that a candidate segment will fall just under one of the algorithm's thresholds.
Am I related to my spouse?
If you are seeing a few regions of identity or half-identity between two unrelated people, that does not necessarily mean that the people are recently related.
It is possible that the relationship is very distant in time and the two people happened to inherit the same piece of DNA from a long-gone ancestor. Finally, it is possible that our algorithm detected a matching segment between two individuals that do not share a recent common ancestor. This is more likely to happen when both people come from a genetically homogeneous population. For example, two people of Ashkenazi Jewish ancestry may often have regions of common ancestry, because many Ashkenazi Jews are actually distant (fifth or sixth) cousins, sharing multiple common ancestors many generations ago.
As scientific knowledge improves and our database grows we will be able to provide increasingly precise and accurate estimates.