Race as most people understand it, a set of distinct biological categories that neatly divide humanity, has no basis in genetics. Human DNA is more than 99.9% identical across all people, and the 0.1% that varies does not sort into the discrete groups we call races. Race is, however, very real as a social and political system that shapes lives, health outcomes, and opportunities. So the short answer is: race is a myth in biology but a powerful reality in society.
What Genetics Actually Shows
The Human Genome Project confirmed that all humans share more than 99.9% of their DNA. The remaining fraction, roughly 3 million base pairs out of nearly 3 billion, accounts for all the visible and invisible differences between any two people on Earth. The critical finding is how that variation is distributed.
In 1972, geneticist Richard Lewontin analyzed protein markers and found that about 85% of all human genetic diversity exists between individuals within the same population. Only about 6 to 7% of variation fell between what were traditionally called racial groups. Later studies using far more sophisticated tools, including microsatellite and restriction fragment data, confirmed this pattern with remarkable consistency: roughly 84% of molecular diversity sits within populations, and differences among continental groups account for only 8 to 12% of the total. Two randomly chosen people from the same village in Nigeria can be more genetically different from each other than either is from a person in Norway.
This means there is more genetic variation within any self-identified racial group than between them. Genetic isolation, sharp boundaries, and distinct evolutionary lineages separating human “races” simply do not exist.
Why Humans Vary Gradually, Not in Categories
Human genetic variation follows what scientists call clines: gradual shifts in trait frequencies across geography. Skin color, for example, doesn’t jump from one shade to another at a border. It shifts incrementally along gradients shaped by UV exposure over thousands of years. The same is true for most genetically influenced traits. They blend smoothly across regions rather than clustering into packages that match racial labels.
There is a nuance here. When researchers use clustering algorithms on large genetic datasets without any geographic labels, people do sort into broad groups that loosely correspond to continental regions: sub-Saharan Africa, Europe and Western Asia, East Asia, Oceania, and the Americas. These clusters are real features of human genetic structure, driven by small jumps in genetic distance across natural barriers like oceans, the Himalayas, and the Sahara Desert. Within each cluster, genetic distance increases smoothly with geographic distance, consistent with clinal variation. But between clusters, genetic distances are slightly larger than geography alone would predict.
These continental clusters, however, are not races. They overlap extensively, many populations show partial membership in multiple clusters, and the boundaries are fuzzy rather than fixed. The clusters reflect migration history and geographic barriers, not the rigid five-race model taught for centuries.
Where the Race Concept Came From
The scientific concept of race traces back to 18th-century European taxonomy. Johann Friedrich Blumenbach, a German physician, published his classification of human “varieties” in 1775. His first edition described four geographically defined groups; by the second edition, he had expanded it to five, eventually labeling them Caucasian, Mongolian, Ethiopian, American, and Malay. Blumenbach himself insisted that humanity was one species and warned against treating his categories as sharp divisions. He acknowledged gradations between groups and argued for the equal potential of all people.
But his system was already contaminated by bias. He declared a Georgian woman’s skull the “most handsome,” speculated that white skin was the original human color because “it is easy to change from white to brown but not vice versa,” and placed Europeans at the center of his hierarchy. These aesthetic judgments, not genetic evidence, shaped the framework. Over time, his nuanced warnings were ignored, and the simplified five-race model became a tool for ranking human groups, justifying slavery, colonialism, and exclusion. As one review put it, social scientists have long understood race to be a social category invented to justify slavery, and evolutionary biologists know that socially constructed racial categories do not align with genetic reality.
Ancestry Is Not the Same as Race
If race isn’t biologically valid, what about ancestry? The distinction matters. Ancestry describes the geographic and genealogical history of your actual forebears. It’s specific, traceable, and genetically meaningful. You might have ancestry from West Africa, the British Isles, and Southeast Asia, each contributing identifiable genetic variants. Race, by contrast, collapses all of that into a single label like “Black” or “White” based primarily on appearance and social context.
The problem with using race as a stand-in for ancestry is that it erases the enormous diversity within any racial category. Labeling someone as having “African ancestry” and treating that as a genetic category implies a single mutation history shared by everyone from Morocco to Mozambique, regardless of whether they share 1% or 99% of their ancestry from the continent. One researcher compared this to the genetic version of the “one-drop rule,” where any trace of African heritage assigns a person to a single category regardless of their actual genetic makeup. Self-identification, which is the standard in social science, is useful for understanding cultural identity but cannot serve as a proxy for DNA.
Real Consequences in Medicine
Race may not be genetic, but it has been embedded in medical practice in ways that cause measurable harm. One clear example: for years, kidney function tests used a race-based correction factor. The standard equations for estimating how well kidneys filter waste automatically adjusted the result upward for Black patients by 16 to 18%, based on the assumption that Black people have higher average muscle mass. That assumption was never supported by credible research.
The consequences were concrete. A higher estimated kidney function score means the disease looks less severe on paper. Black patients were less likely to be referred for specialist care, less likely to be placed on transplant waiting lists, and more likely to have their kidney disease undertreated. When a national task force of 97 experts across 21 states and seven countries reviewed the evidence, they recommended removing the race coefficient entirely. A subsequent study found that doing so reclassified one third of Black patients to a more severe stage of chronic kidney disease, meaning they had been systematically undertreated.
The same pattern appears in drug treatment. Pharmacogenomic testing, which examines specific gene variants that affect how your body processes medications, provides a far more accurate picture than racial labels ever could. If all patients had access to individual genetic testing, race and ethnicity would be irrelevant to treatment decisions. The variation that matters for drug metabolism exists within every population, not between racial categories.
What Forensic Science Shows (and Doesn’t)
Forensic anthropologists can estimate a person’s likely continental ancestry from skeletal remains with roughly 91% accuracy, which might seem to support the idea of biological races. But what this actually demonstrates is that geographic ancestry leaves traces in bone structure, not that those traces sort neatly into racial categories. The estimates work best when comparing individuals from well-separated geographic regions. They become far less reliable for people with mixed ancestry or from populations that don’t fit standard reference groups. Accuracy has improved over time largely because analysts now use metric measurements and statistical tools rather than visual classification alone.
These skeletal differences reflect the same clinal variation seen in genetics: real, measurable geographic patterns that don’t map onto social racial categories in any clean way.
A Social Reality With Biological Effects
Dismissing race as a biological myth does not mean dismissing its impact. Race, as a social system, produces real biological outcomes. Chronic stress from discrimination raises cortisol levels, damages cardiovascular health, and shortens life expectancy. Residential segregation concentrates environmental hazards like air pollution and lead exposure in communities of color. Unequal access to healthcare, nutrition, and safe housing creates health disparities that can look genetic on the surface but are driven by social determinants.
The genome.gov definition captures this precisely: race is a social construct used to group people, constructed as a hierarchical system to identify, distinguish, and marginalize some groups. It is political, social, and fluid. It changes across time and geography. A person classified as one race in Brazil may be classified differently in the United States. These are not the characteristics of a biological category.
The effects of racism, not race itself, are responsible for disparities in health outcomes. Understanding this distinction is the difference between looking for broken genes in marginalized communities and looking at the systems that break people’s health.