For decades, textbooks and public-health campaigns repeated a striking claim: bacterial cells in the human body outnumber human cells by a factor of ten to one. That figure, repeated so often it became biological common knowledge, turns out to be wrong. Revised cell-by-cell accounting for a 70-kilogram adult male now places the ratio at roughly one to one, with about 3.8 trillion bacterial cells living alongside about 3.0 trillion human cells. The correction reshapes how scientists talk about the microbiome, and it raises pointed questions about how an unverified number persisted for more than four decades in peer-reviewed literature and government research programs.
Why the revised bacterial-to-human cell ratio changes the conversation
The old ten-to-one figure did real work. It shaped grant applications, framed media coverage, and gave the microbiome field an easy hook: you are more microbe than human. Researchers at the Weizmann Institute of Science and the Hospital for Sick Children in Toronto traced the claim back to a 1972 estimate by Thomas Luckey, according to an audit published in Cell. Luckey’s calculation relied on rough assumptions about colon volume and bacterial density that were never re-examined with modern tools before being cited thousands of times.
The NIH Human Microbiome Project published coordinated results in 2012, cataloging microbial diversity across hundreds of healthy volunteers. Even at that stage, project-associated literature still echoed the legacy claim. A peer-reviewed overview of the HMP’s goals, for instance, repeated the assertion that there are “at least 10 times more bacteria than human cells,” attributing it to D.C. Savage’s 1977 review of gastrointestinal microbial ecology. Savage’s discussion of gut colonization, available through an Annual Review of Microbiology article, helped cement the idea that microbes vastly outnumber their human hosts, even though the underlying numbers were approximate.
Whether the tenfold figure traces primarily to Luckey in 1972 or to Savage in 1977 remains a point of bibliographic dispute, but both citations relied on order-of-magnitude reasoning rather than direct measurement. Over time, a rough back-of-the-envelope estimate hardened into a “fact” that shaped how scientists, journalists, and clinicians described the human body. The practical consequence is straightforward. When researchers describe immune training, metabolic signaling, or disease susceptibility in terms of a vast bacterial majority, they are building on a premise that overstates the numerical imbalance by roughly tenfold. That does not make the microbiome less important, but it does mean the framing needs recalibration.
In public communication, the ten-to-one slogan encouraged a kind of microbial exceptionalism: if bacteria so thoroughly outnumber human cells, then perhaps they are the real drivers of physiology and behavior. Advocates of microbiome-centered medicine used the ratio to argue that altering gut communities might unlock sweeping health benefits. The newer estimate, by contrast, suggests a more balanced picture in which microbial and human cells coexist in comparable numbers. The emphasis shifts from numerical dominance to functional interplay: what matters is less how many cells there are and more what they are doing.
How Sender, Fuchs, and Milo recounted the cells
The revised estimate comes from a 2016 analysis by Ron Sender, Shai Fuchs, and Ron Milo, published in PLOS Biology. For a 70-kilogram reference man, the team calculated approximately 3.8 times 10 to the 13th bacterial cells and approximately 3.0 times 10 to the 13th human cells, yielding a ratio close to 1.3 to 1. The dominant bacterial compartment is the colon, which houses the vast majority of the body’s microbes. On the human side, red blood cells account for roughly 70 percent of the total cell count, a detail that earlier estimates often overlooked or undercounted.
Sender and colleagues built their bacterial tally by combining measurements of colon content volume with typical bacterial densities reported in the literature. They also accounted for smaller contributions from other body sites, such as the skin and oral cavity, which host dense microbial communities but represent a relatively small fraction of the total. Their approach did not involve counting bacteria directly in living people; instead, it synthesized the best available anatomical and microbiological data into a coherent whole-body estimate.
On the human side of the ledger, the team broke the body into major cell types and organs, then used known volumes, masses, and cell densities to derive counts. This method highlighted how numerically dominant small cells like erythrocytes are compared with larger but fewer cells such as muscle fibers or neurons. Earlier narratives that implicitly treated all tissues as contributing equally to cell number missed this skew, making it easier to understate the human total and inflate the relative weight of bacteria.
An independent estimate of total human cells, published in the Annals of Human Biology by Bianconi and colleagues, arrived at a figure of approximately 3.7 times 10 to the 13th. That number is higher than the Sender team’s 3.0 times 10 to the 13th, and the gap reflects different choices about which cell types to include and how to estimate organ-level densities. Both figures, however, land in the same order of magnitude, and both confirm that the bacterial side of the ledger is not ten times larger. Instead, they suggest a narrow range in which human and microbial cells are peers, not distant rivals.
The hypothesis that individuals whose bacterial-to-human cell ratios skew above 1.5 to 1 would show measurably higher short-chain fatty acid output in stool metabolomics, independent of diet, is testable in principle. Short-chain fatty acids such as butyrate and propionate are produced almost entirely by colonic bacteria, and a higher absolute bacterial load in the colon could plausibly increase their output. No study in the available evidence base, however, has directly measured whole-body bacterial-to-human cell ratios in living subjects and correlated those ratios with metabolomic profiles. The hypothesis remains unconfirmed, and any claims linking the revised ratio to specific metabolic outcomes should be treated as speculative.
Gaps in the evidence and what to watch next
Several limitations constrain how far the revised ratio can be applied. The Sender team’s bacterial estimate relies on extrapolations from colon volume and bacterial density measurements, not on whole-body counts taken from living individuals. The reference figure applies to a 70-kilogram adult male; variation across sex, age, body composition, and geography has not been systematically characterized in the same framework. A bowel movement alone can shift the ratio temporarily, as the authors themselves noted, because a large fraction of the body’s bacteria reside in fecal matter transiting the colon.
These caveats matter for interpretation. If bacterial numbers can swing noticeably over the course of a day, then the idea of a single stable “ratio” becomes more of a statistical convenience than a personal metric. Likewise, people with very low or very high body mass, altered gut anatomy, or chronic gastrointestinal disease may fall outside the assumptions built into the reference model. The current estimates are best understood as population-level benchmarks, not individualized diagnostics.
Another gap involves function. Cell counts say nothing about which microbial species are present, how active they are, or what molecules they produce. Two people could share a similar bacterial-to-human cell ratio yet differ dramatically in immune tone or metabolic risk because their microbial communities carry different genes and respond differently to diet, drugs, and infection. Future studies that integrate cell-number estimates with genomic, transcriptomic, and metabolomic data will be needed to connect the revised ratio to concrete health outcomes.
Still, the recalibration has immediate implications for scientific communication. Grant proposals, review articles, and educational materials that lean on the ten-to-one slogan will need updating. More broadly, the episode is a case study in how convenient numerical narratives can persist long after their empirical basis has gone stale. As microbiome research matures, the field may benefit from fewer sweeping ratios and more attention to carefully bounded, testable claims about how microbes and human cells share the same body.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
