How closely related are all humans?

How similar is the DNA of any two people?

Your genome is the full set of DNA instructions in your cells. It is written in four chemical letters: A, T, C and G. One copy has about 3 billion letters, and most cells carry two copies, one from each parent. That makes about 6 billion letters in all, according to NHGRI’s fact sheet on human genomic variation.

Compared with a reference genome, a typical person differs at about 5 million single letters. On top of that come about 600,000 small insertions or deletions, and about 25,000 structural variants: stretches of 50 or more letters that are added, missing, flipped or moved. Together these touch about 27 million letters, or about 0.4% of the genome. That is why NHGRI says any two genomes are, on average, about 99.6% identical. The often-quoted 99.9% figure counts only single-letter differences, which show up about once every 1,300 letters. NHGRI calls that an over-simplification. Either way, we are nearly identical.

We also have less genetic variety than many other species. A pair of human chromosomes differs at only about 1 in 1,000 sites, compared with about 3 in 1,000 in chimpanzees, according to a 2023 National Academies report. The Smithsonian adds that one chimpanzee subspecies, living only in central Africa, has more genetic diversity than all humans worldwide.

Where does our shared family tree lead?

Back to Africa. The National Academies report sums it up: anatomically modern humans arose somewhere in Africa about 300,000 years ago, and their descendants spread across much of the rest of the world within the past 100,000 years. Fossils from Jebel Irhoud, in Morocco, are dated to about 315,000 years ago, give or take 34,000 years. The team that described them in 2017 concluded that the processes that produced our species involved the whole African continent.

DNA agrees. The 1000 Genomes Project read the genomes of 2,504 people from 26 populations. People from populations with African ancestry carried the most genetic variants, as the out-of-Africa model predicts. Most common variants turned up all over the world, while rarer ones tended to stay within closely related populations. The story is not perfectly tidy: Neanderthals contributed up to 1 to 4% of the genomes of people whose ancestors lived outside Africa, and a 2020 study found Neanderthal DNA in African populations too, a sign that genes flowed back and forth over time. Even so, the Smithsonian says, most of our genome still comes from Africa.

You also have two kinds of family tree. Your genealogical tree holds everyone you descend from: parents, grandparents, great-grandparents and so on. Your genetic tree holds only the ancestors whose DNA actually reached you. Each parent passes on only half of their DNA, reshuffled, and most of your distant ancestors left no DNA in you, the National Academies explain.

Genealogical trees overlap fast. Each step back doubles the places in the tree: 2 parents, 4 grandparents, 8 great-grandparents. Thirty generations back there are more than a billion places. Because there have only ever been so many people, the same ancestors fill many places, and different people’s trees soon merge. In 2004, Douglas Rohde, Steve Olson and Joseph Chang built two models of this. Both included real-world limits: people tend to choose partners from their own group, and groups far apart are partly cut off. In both models, the most recent common ancestor of everyone alive today lived just a few thousand years ago. Go back a few thousand years more, and in the models everyone alive today has exactly the same set of ancestors.

These are models, not measurements, so the dates are estimates. The National Academies call the result likely: by several thousand years ago, all modern humans probably share all their genealogical ancestors. Shared genetic ancestors are much older: for any one stretch of DNA, the most recent ancestor that everyone shares is typically far more ancient (Matsen and Evans, 2008). Our answers on how life on Earth began and on whether evolution would turn out the same if replayed trace the longer story behind us.

Is most variation between groups, or within them?

Within them. In 1972, R. C. Lewontin published a study of how human genetic diversity is divided among groups. The split often quoted in this debate, as A. W. F. Edwards summed it up in 2003, is about 85% of the variation within populations and 15% between populations or ethnic groups. Later studies found the same pattern. A 2002 study of 1,056 people from 52 populations found that differences among people within a population made up 93 to 95% of the genetic variation, and differences among major groups only 3 to 5%. The Smithsonian’s summary: in general, 5 to 15% of human genetic variation lies between large groups living on different continents.

Gene variants also tend to change in frequency gradually with distance, with slightly bigger jumps across barriers such as oceans or mountains. The National Academies note that these geographic boundaries do not line up with racial groupings. NHGRI’s glossary calls race a social construct used to group people, often by physical appearance, social factors and cultural background, and it notes that there is more genetic variation within self-identified racial groups than between them. Large genome projects, the National Academies report says, confirmed a poor correspondence between racial groups and the way human genetic variation is spread. Its 2023 recommendations say researchers should not use race as a stand-in for genetic variation.

Looks can mislead. Skin color, for example, closely tracks the strength of sunlight where a person’s ancestors lived, the Smithsonian explains. Darker skin protects against strong ultraviolet light near the equator, and lighter skin lets the body make enough vitamin D where sunlight is weaker.

What is the strongest objection?

If race is a social label, how can a DNA test tell where your ancestors lived? It is a fair question. In 2003, Edwards argued that Lewontin’s conclusion, that the genetic data do not justify dividing humans into such groups, was unwarranted. Most of the information that tells populations apart, he wrote, lies in the way many markers vary together, not in any single marker. The 2002 study is an example. Without using any information about where people came from, the researchers found six main genetic clusters, and five of them matched major geographic regions.

Both findings can be true at once, a 2007 study explained. The share of variation between populations is modest, and two people from different populations can be more alike than two people from the same one. Yet with enough data, people can be sorted into populations accurately. The National Academies add that the labels given to such groups, such as a region or an ethnicity, come from outside genetics, and that the underlying variation changes continuously from place to place. So genetic ancestry is real and can be estimated, but it changes gradually across geography. It is not a set of sharp boxes, and it does not map neatly onto race.

What do people often get wrong?

“If we are 99.9% the same, genes cannot matter.” Small differences can matter. NHGRI notes that some variants influence traits like eye color and height, and some directly cause diseases such as cystic fibrosis and sickle cell disease. Being nearly identical and being unique are both true.

“If we all share ancestors from a few thousand years ago, we all carry their DNA.” Not necessarily. Most of a person’s distant genealogical ancestors passed no DNA to them, the National Academies explain, and which ancestors did pass on DNA changes from one part of the genome to another.

“People from one continent share one genetic type.” The 1000 Genomes Project found the most genetic variants in people from populations with African ancestry. A single label for a whole continent hides a great deal of variety. The National Academies report cites researchers who call continental labels such as European, African and Asian very broad.

How can you check this yourself?

• Read NHGRI’s fact sheet. It explains where the 99.9% and 99.6% figures come from.
• Ask what was counted. When you see a percentage, check whether it counts single letters or every kind of difference.
• Read how ancestry estimates work. Methods measure how similar your DNA is to other people’s, then borrow geographic or other labels from genetically similar people, the National Academies explain. The labels come from outside genetics.
• Visit the Smithsonian’s Human Origins pages. They show the fossil and genetic evidence for our African origin.
• Check the date. New fossils and ancient DNA keep refining the timeline, so notice when a source was written.

The answer stays the same. Every person alive belongs to one young species, and we are far more alike than different. In the National Academies’ words, “modern humans are all embedded within the same enormous family tree, with some members more closely related than others.”