The honest answer is that the world’s most prominent scientific and regulatory bodies have looked at largely the same pile of evidence and reached different conclusions. The International Agency for Research on Cancer (IARC), part of the World Health Organization, classified glyphosate as “probably carcinogenic to humans” in 2015. The U.S. Environmental Protection Agency, the European Food Safety Authority, and several other national regulators concluded that it is not likely to cause cancer at real-world exposure levels. That split is not just politics or corruption, though those accusations fly in both directions. It reflects genuine methodological disagreements about which studies count, how to weigh animal data against human data, and whether the chemical itself or the commercial products containing it should be the focus of evaluation.
Why the Major Agencies Reached Different Conclusions
The disagreement between IARC and regulatory agencies like the EPA and EFSA is one of the most studied disputes in modern toxicology. A detailed review in Archives of Toxicology found that part of the divergence stems from different data sets: IARC relied heavily on published, peer-reviewed literature, while regulatory agencies also incorporated unpublished industry-submitted studies from the companies that manufacture glyphosate. But even beyond data selection, the two sides used different evaluation methods to judge the same evidence.1PubMed Central. Glyphosate toxicity and carcinogenicity: a review of the scientific basis of the European Union assessment and its differences with IARC
IARC evaluates hazard: can a substance cause cancer under any realistic circumstance, at any dose? Regulatory agencies evaluate risk: is the substance likely to cause cancer at the doses people actually encounter? Those are fundamentally different questions. A substance can be a hazard (capable of causing harm in theory) without posing meaningful risk at everyday exposure levels. This distinction matters, because both sides can be technically correct while appearing to flatly contradict each other. Critics of the regulatory process, however, argue that the heavy reliance on unpublished industry data, combined with documented cases of industry influence on regulators, undermines confidence in the “no risk” conclusion.2PubMed Central. Glyphosate, Roundup and the Failures of Regulatory Assessment
What Human Epidemiology Shows
The largest single study tracking glyphosate users is the Agricultural Health Study (AHS), a prospective cohort of over 50,000 licensed pesticide applicators in the United States. The study followed farmers for years and found no association between glyphosate use and any solid tumors or lymphoid malignancies overall, including non-Hodgkin lymphoma (NHL) and its subtypes.3PubMed Central. Glyphosate Use and Cancer Incidence in the Agricultural Health Study That is a substantial piece of reassuring evidence, because prospective cohort studies are generally more reliable than case-control studies. They track people forward in time rather than asking cancer patients to remember what chemicals they used years ago.
But the picture changes when researchers pool data from multiple studies and focus specifically on people with the highest cumulative exposure. A 2019 meta-analysis incorporating the updated AHS data found a statistically significant 41% increased risk of NHL among people with the highest glyphosate exposure over time.4PubMed Central. Exposure to Glyphosate-Based Herbicides and Risk for Non-Hodgkin Lymphoma: A Meta-Analysis and Supporting Evidence That finding drove much of the concern that led to high-profile lawsuits against Monsanto (now Bayer).
Not every meta-analysis agrees. An updated analysis published in 2021 found no overall association between ever having used glyphosate and NHL risk, though it could not rule out an association with one specific subtype called diffuse large B-cell lymphoma.5PubMed Central. Exposure to glyphosate and risk of non-Hodgkin lymphoma: an updated meta-analysis Another meta-analysis in Cancer Causes and Control found that the strength of the association depended heavily on how much lag time was assumed between exposure and cancer diagnosis. When analyses used a 20-year exposure lag, the association was significant. Without a lag, the summary estimate was not statistically significant.6PubMed. On recent meta-analyses of exposure to glyphosate and risk of non-Hodgkin’s lymphoma in humans
An earlier systematic review found a borderline-significant association between any glyphosate use and both NHL and multiple myeloma, but noted the evidence was constrained by few studies, crude exposure measurements, and the possibility of confounding from other pesticides. The authors concluded that a causal relationship had not been established.7PubMed Central. Systematic review and meta-analysis of glyphosate exposure and risk of lymphohematopoietic cancers This is worth noting because farmers who use glyphosate rarely use only glyphosate. Disentangling its effects from dozens of other chemical exposures is an ongoing challenge in agricultural epidemiology.
What Animal Studies Show, and Why Experts Disagree About Them
Animal bioassays are the other major line of cancer evidence, and they are just as contested as the human data. A comprehensive reanalysis of all available chronic-exposure rodent studies identified 37 statistically significant tumor findings across mice and rats. The strongest evidence pointed to hemangiosarcomas, kidney tumors, and malignant lymphomas in certain mouse strains, as well as kidney, liver, and skin tumors in rats.8PubMed Central. A comprehensive analysis of the animal carcinogenicity data for glyphosate from chronic exposure rodent carcinogenicity studies
An industry-convened expert panel reviewed much of the same data and reached the opposite conclusion, finding that the tumor increases either lacked statistical significance, fell within historical control ranges, did not follow a dose-response pattern, or did not progress from benign to malignant tumors.9PubMed. Glyphosate rodent carcinogenicity bioassay expert panel review The weight-of-evidence assessment from that panel was that glyphosate is not a carcinogen in laboratory animals.
A 2025 study complicated the picture further. Researchers exposed Sprague-Dawley rats to glyphosate and glyphosate-based herbicides from prenatal life onward and observed statistically significant increases in tumors at multiple sites, including leukemia, liver, thyroid, nervous system, and other organs. Many of these tumors were types considered rare in the rat strain used, and about 40% of leukemia deaths in treated groups occurred before one year of age, suggesting an unusually early onset.10PubMed Central. Carcinogenic effects of long-term exposure from prenatal life to glyphosate and glyphosate-based herbicides in Sprague-Dawley rats Lifetime exposures starting before birth have not been the standard protocol in most earlier regulatory studies, so this kind of design may reveal effects that conventional two-year bioassays miss.
The Formulation Problem
One of the most underappreciated complications in the glyphosate debate is that commercial herbicide products are not just glyphosate. Products like Roundup contain surfactants and other additives that help the active ingredient penetrate plant cells. Multiple studies have found that these complete formulations are substantially more toxic to human cells than glyphosate alone.
A study comparing the cytotoxicity of Roundup Classic’s components found an unambiguous hierarchy: the surfactant POE-15 was the most toxic, the full Roundup formulation was considerably less so but still problematic, and glyphosate by itself was far less harmful.11PubMed Central. Cytotoxic effects of Roundup Classic and its components on NE-4C and MC3T3-E1 cell lines determined by biochemical and flow cytometric assays Research on cardiac cells found that Roundup reduced mitochondrial respiration by nearly half compared to controls, while glyphosate alone had no significant effect.12PubMed Central. Effects of Glyphosate and Roundup® Herbicides on Cardiac and H9c2 Cells’ Mitochondrial Respiration and Oxidative Stress DNA damage testing on human cell lines showed that certain commercial formulations like Roundup Mega caused significant genotoxic effects at concentrations where individual surfactants did not.13PubMed Central. Comparative cytotoxicity and genotoxicity of commercial glyphosate-based herbicide formulations and co-formulants in human leukocyte and hepatocyte cell lines
This matters because regulatory agencies have historically evaluated glyphosate as an isolated chemical, while IARC considered evidence on both the active ingredient and the formulated products. A review in the Journal of Xenobiotics specifically highlighted this gap, noting the growing evidence for toxicity from both glyphosate metabolites and the surfactants present in commercial herbicides.14PubMed Central. Glyphosate vs. Glyphosate-Based Herbicides Exposure: A Review on Their Toxicity If the surfactants are doing much of the damage, then studying glyphosate in isolation and declaring it safe may not tell the full story about what actually happens when a farmer mixes and sprays a jug of Roundup.
How Much Glyphosate People Are Actually Exposed To
For all the debate about whether glyphosate can cause cancer, the practical question many people care about is: how much of it is actually getting into their bodies? Biomonitoring studies consistently find detectable glyphosate in most people’s urine, including those who never touch a herbicide sprayer. A large study of U.S. farmers and non-farmers found glyphosate in the urine of 81% of non-farming controls, though at lower concentrations than in farmers who had recently applied it.15PubMed Central. Urinary biomonitoring of glyphosate exposure among male farmers and nonfarmers in the Biomarkers of Exposure and Effect in Agriculture (BEEA) study
Occupational exposure is significantly higher. A review of eight studies covering over 400 occupationally exposed workers found average urinary glyphosate levels ranging from about 0.26 to 73.5 micrograms per liter, while environmental exposure levels ranged from 0.16 to 7.6 micrograms per liter. Both studies that tracked trends over time found an increasing proportion of individuals with detectable glyphosate.16PubMed Central. The evidence of human exposure to glyphosate: a review An Italian study of vineyard workers found that glyphosate was undetectable before a single application day, but present in every worker’s urine the day after, with some workers reaching nearly 50 micrograms per liter.17PubMed Central. Short occupational exposure to glyphosate and its biomonitoring via urinary levels of glyphosate and metabolite AMPA (Amino-MethylPhosphonic acid), in Italian vineyard workers
For people living near agricultural fields, spray drift is an additional route. A study of pregnant people in southern Idaho found evidence that proximity to agricultural operations during spray season increased glyphosate exposure, whether through direct drift, contaminated soil, or dust particles entering residences.18PubMed Central. The Effect of Pesticide Spray Season and Residential Proximity to Agriculture on Glyphosate Exposure among Pregnant People in Southern Idaho, 2021
Dietary Residues and Safety Margins
For most people who do not work in agriculture or live next to farms, food is the main source of glyphosate exposure. Glyphosate is commonly detected in grain products, legumes, fruits, and vegetables. A Canadian government survey of nearly 8,000 food samples found detectable glyphosate residues in about 42% of them. The vast majority, roughly 99.4%, complied with Canadian regulatory limits, and the agency determined there was no long-term health risk from the levels found.19Journal of Agricultural and Food Chemistry. Analysis of Glyphosate Residues in Foods from the Canadian Retail Markets between 2015 and 2017 A broader review of international food monitoring found glyphosate prevalence ranging from 8% to 100% depending on the food type tested, with only about 1% of samples exceeding maximum residue limits.20Journal of Agriculture and Food Research. Glyphosate in food: A narrative review
Even at the high end of dietary intake, glyphosate levels from food appear to remain well below the EPA’s chronic reference dose, which is set at 0.1 milligrams per kilogram of body weight per day. An analysis of dietary exposure across three common U.S. eating patterns found that even the 95th percentile of intake was comfortably below that threshold. Whether that threshold is set at the right level, of course, is its own debate. Some researchers argue that current safety margins were established without adequate consideration of long-term low-dose effects and the added toxicity of commercial formulations.
Endocrine Disruption and Cellular Effects
Cancer is not the only health concern with glyphosate. Research has found that glyphosate-based formulations can disrupt endocrine function in human cell lines. One study reported that several commercial formulations disrupted androgen receptor activity starting at 0.5 parts per million, well below typical agricultural application concentrations. Estrogen receptor activity and aromatase enzyme function were also affected at slightly higher concentrations. The effects were more dependent on the specific formulation than on the concentration of glyphosate itself, reinforcing the concern about co-formulants.21PubMed. Glyphosate-based herbicides are toxic and endocrine disruptors in human cell lines
Kidney effects have also drawn attention. A study combining human epidemiological data with zebrafish experiments found a negative association between glyphosate exposure and renal function in community-dwelling adults. The zebrafish experiments showed kidney damage including cystic dilation at high glyphosate concentrations, along with disrupted metabolic pathways.22PubMed. Effect of glyphosate on renal function: A study integrating epidemiological and experimental evidence Separate fish studies found that chronic low-level glyphosate exposure caused extensive damage to kidney tubule cells, with vacuolation indicating abnormal accumulation of material inside cells.23Toxicological Sciences. Chronic exposure to low levels of glyphosate and metals induces kidney dysfunction These non-cancer effects may matter more to public health than the cancer question for people with lower-level but chronic exposures.
The Gut Microbiome Question
Glyphosate kills plants by blocking an enzyme called EPSPS in the shikimate pathway, which plants use to make certain amino acids. Humans do not have this pathway, which is one reason glyphosate has long been considered relatively safe for people. But many gut bacteria do have it. A survey of over 700 bacterial strains from the Human Microbiome Project found that roughly 55% are potentially sensitive to glyphosate based on the structure of their EPSPS enzyme. Sensitive genera include some considered beneficial, like Bifidobacterium, while several resistant genera have been linked to conditions like irritable bowel syndrome.24PubMed Central. Does Glyphosate Affect the Human Microbiota?
That sounds alarming on paper, but other research introduces important caveats. Computational modeling of the gut microbiome found that most gut bacteria do not possess a complete shikimate pathway and that this pathway is mostly transcriptionally inactive in the human gut. The implication is that most gut bacteria may be getting their aromatic amino acids from their environment rather than making them, making them relatively resistant to glyphosate’s mechanism of action.25PubMed Central. Computational modelling provides insight into the effects of glyphosate on the shikimate pathway in the human gut microbiome Whether dietary-level glyphosate exposure is enough to meaningfully shift the gut microbiome in living humans remains an open question. Animal studies have shown that low-dose glyphosate-based herbicide exposure can alter metabolites related to one-carbon metabolism, a pathway that depends on B vitamins partially produced by gut bacteria, but translating those findings to typical human dietary exposure levels is speculative at this point.
Epigenetic Effects Across Generations
Some of the most provocative glyphosate research involves epigenetics, the study of how chemical exposures can alter gene expression without changing DNA itself. In a rat study, pregnant females were exposed to glyphosate during a critical developmental window. The researchers then examined the great-grandchildren of those rats, animals that had never been directly exposed. These third-generation offspring showed increased rates of prostate disease, kidney disease, and obesity, along with distinct patterns of altered DNA methylation and histone retention in their sperm.26PubMed Central. Epigenome-wide association study for glyphosate induced transgenerational sperm DNA methylation and histone retention epigenetic biomarkers for disease
These transgenerational findings are still preliminary and come from animal models using exposure levels that may not reflect human experience. But they raise a type of concern that conventional toxicology studies, which typically look at the exposed animal and stop, are not designed to detect. If glyphosate can leave molecular marks that persist across generations without continued exposure, the implications extend well beyond the current cancer debate.
Environmental Persistence and Its Metabolite AMPA
Glyphosate was long marketed as a chemical that breaks down quickly in soil, and under warm, moist conditions that can be true: half-lives as short as 1.5 days have been measured. But in cold or dry conditions, the half-life can stretch past 50 days, and the time for 90% degradation can exceed 280 days.27PubMed. Persistence of glyphosate and aminomethylphosphonic acid in loess soil under different combinations of temperature, soil moisture and light/darkness In temperate climates with cold, dry winters, repeated applications can lead to accumulation.
Glyphosate’s primary breakdown product, aminomethylphosphonic acid (AMPA), persists in soil even longer than glyphosate itself. A multi-year study in Quebec found that while glyphosate was no longer detectable in soil samples, AMPA was present in roughly half of them. In some cases, the amount of AMPA in the soil exceeded what could be explained by the previous season’s glyphosate application, suggesting carryover from earlier years.28Agriculture. Multi-Year Pseudo-Persistence, Mobility, and Degradation of Glyphosate and Its Degradation Product (AMPA) in a Gleysol in Quebec (Canada) AMPA has its own toxicological profile that is not yet as well characterized as glyphosate’s, and its accumulation in agricultural soils is an active area of investigation.
What Happens If Glyphosate Is Restricted
The European Union has repeatedly revisited whether to renew glyphosate’s authorization, and some countries and municipalities have imposed partial bans. This raises a practical question: are the alternatives better? A Belgian study comparing glyphosate-based strategies against chemical and non-chemical alternatives across several agricultural settings found that the trade-offs are real. Replacing glyphosate with other herbicides can shift environmental burdens rather than eliminate them, sometimes increasing climate impact or ecotoxicity in different categories.29Agronomy. Replacing Glyphosate Shifts Environmental Burdens: Trade-Offs Between Ecotoxicity and Climate Impact in Chemical and Non-Chemical Strategies Non-chemical methods like mechanical weeding or cover cropping work for some applications but are labor-intensive and not always practical at scale. The discussion around glyphosate restriction is not as simple as removing a harmful chemical; it requires reckoning with what takes its place and whether that substitution actually reduces total harm to health and ecosystems.