Glyphosate

Glyphosate is the world's most widely used herbicide — the active ingredient in Roundup and hundreds of other weed-killers. Since the mid-1970s it has become woven into modern agriculture, and traces of it now show up in food, water, and the urine of people who have never sprayed a weed in their lives. Few chemicals are argued about more fiercely. On one side, the World Health Organization's cancer agency called it "probably carcinogenic to humans"; on the other, the US Environmental Protection Agency, Europe's food-safety authority, and most national regulators have concluded it is unlikely to cause cancer at the levels people actually encounter. Billions of dollars in jury verdicts have landed on top of the science, making the picture even harder to read. This page tries to lay out what is genuinely known, what is genuinely disputed, and what is simply unresolved — without either the activist alarm or the industry shrug. The honest summary is that glyphosate is not a proven human carcinogen, but it is also not a closed case, and reasonable scientists still disagree about how much the remaining uncertainty should worry you.

Table of Contents

  1. What Glyphosate Is and How It Works
  2. Where Exposure Comes From
  3. The Cancer Classification Controversy
  4. Non-Hodgkin Lymphoma: What the Studies Show
  5. The Gut-Microbiome and Shikimate Hypothesis
  6. Formulations, Surfactants, and "Glyphosate Alone"
  7. Lawsuits and Settlements vs. the Science
  8. Realistic Exposure and Regulatory Limits
  9. How to Reduce Your Exposure
  10. The Honest Bottom Line
  11. Research Papers
  12. Connections
  13. Featured Videos

What Glyphosate Is and How It Works

Glyphosate (chemically, N-(phosphonomethyl)glycine) is a broad-spectrum systemic herbicide, meaning it kills a wide range of plants and moves through them rather than just burning the leaf it lands on. It was discovered by a Monsanto chemist, John Franz, in 1970 and brought to market as Roundup in 1974. When the patent expired in 2000, dozens of manufacturers began making generic versions, and it is now sold under hundreds of brand names worldwide.

What makes glyphosate scientifically interesting — and central to the safety debate — is how it kills plants. It blocks a plant enzyme called EPSP synthase (5-enolpyruvylshikimate-3-phosphate synthase), which is a key step in the shikimate pathway. Plants use that pathway to build three essential amino acids (phenylalanine, tyrosine, and tryptophan). Shut the pathway down and the plant starves of the building blocks it needs, then wilts and dies over one to two weeks.

Here is the point defenders lean on: humans and other animals do not have the shikimate pathway at all. We get those three amino acids from our diet, so we have no EPSP synthase for glyphosate to jam. On its face, that suggests the chemical targets something we simply do not possess — a genuine biological reason to expect low mammalian toxicity, and the reason glyphosate was long considered one of the safer herbicides. The counter-argument, which we will come back to, is that our gut bacteria do have the shikimate pathway, and that "the target isn't in human cells" is not the same as "it can do nothing in a human body." Both statements can be true at once, and that tension sits at the heart of the disagreement.

Where Exposure Comes From

Because glyphosate is used on such an enormous scale — global use rose roughly fifteen-fold between 1996 and 2014, driven largely by genetically engineered "Roundup Ready" crops — low-level exposure is now close to universal in industrialized countries. The main routes are:

The gap between these exposure levels is not a footnote; it is the whole ballgame. A wheat farmer spraying for a career and a child eating a granola bar are separated by a factor of thousands, and evidence about one tells you very little about the other.

The Cancer Classification Controversy

In March 2015, the International Agency for Research on Cancer (IARC), the cancer arm of the World Health Organization, classified glyphosate as Group 2A, "probably carcinogenic to humans." Headlines went around the world. Within months, the US EPA, the European Food Safety Authority (EFSA), the European Chemicals Agency, and regulators in Canada, Australia, Japan, and elsewhere reviewed largely the same evidence and reached the opposite practical conclusion: glyphosate is unlikely to be carcinogenic to humans at realistic exposures. How can serious scientists look at the same chemical and disagree so completely? The reasons are real and worth understanding, because "who's lying?" is the wrong question.

Hazard versus risk

This is the single most important distinction on the entire page. IARC assesses hazard — can this substance cause cancer under any circumstances, at any dose? It deliberately does not ask how much exposure it takes. Regulators like EPA and EFSA assess risk — does this substance cause harm at the doses people are actually exposed to? By IARC's hazard yardstick, processed meat, alcoholic beverages, and sunlight are also carcinogens; being on the list says a substance can cause cancer, not that a given exposure will. So IARC saying "probably carcinogenic" and EPA saying "unlikely at real-world doses" are not necessarily contradictory. They are answers to two different questions that the public hears as one.

Which studies each side weighed

IARC leaned heavily on published, peer-reviewed literature and on studies of highly exposed applicators, plus animal and mechanistic (cell and DNA-damage) evidence it judged to show carcinogenic potential. Regulators additionally relied on a large body of unpublished industry-submitted toxicology studies, which they are legally entitled to see and IARC generally does not use. Critics of the regulators argue this reliance on company data is a conflict; defenders argue those studies are conducted to strict standardized protocols and that excluding them throws away the largest and most rigorous animal dataset available. Both concerns are legitimate.

Glyphosate alone versus the full formulation

IARC's review considered "glyphosate" broadly, including glyphosate-based formulations. Regulators, by design, assess the pure active ingredient for the cancer classification, evaluating the added surfactants and other ingredients separately. As we will see, some of those other ingredients may be more biologically active than glyphosate itself — so testing the active ingredient in isolation may not capture what a bottle of Roundup does. This narrows what "glyphosate is safe" can honestly claim.

None of this makes one side dishonest. It reflects two well-established scientific frameworks that were built to answer different questions, applied to a chemical caught in the middle.

Non-Hodgkin Lymphoma: What the Studies Show

The specific cancer at the center of the argument is non-Hodgkin lymphoma (NHL), a cancer of the immune system's white blood cells. The human evidence points in two directions depending on which study you trust most, and honest readers should sit with both.

The Agricultural Health Study (points toward "no clear link")

The Agricultural Health Study (AHS) is the largest and most rigorous prospective cohort on this question — more than 54,000 licensed pesticide applicators in Iowa and North Carolina followed for years, with exposure recorded before anyone got sick (which avoids the memory bias that plagues after-the-fact studies). Its updated 2018 analysis found no statistically significant association between glyphosate use and non-Hodgkin lymphoma, or with most other cancers. This is the strongest single piece of evidence on the reassuring side, and it studied exactly the high-exposure group where a signal should be easiest to see.

The 2019 meta-analysis (points toward "a real signal in the most exposed")

A widely cited 2019 meta-analysis pooled several studies and reported a roughly 41% increased risk of NHL among the most highly exposed individuals. It is influential and was written partly by scientists who had advised IARC. But it has a genuine limitation that its critics press: it emphasized the highest-exposure subgroup across studies, a choice that can amplify a signal, and much of its weight comes from case-control studies that ask people to recall past exposure — a design more prone to bias than the AHS cohort.

How to hold both honestly

The fair reading is not "one study is right and the other is propaganda." It is that if glyphosate raises NHL risk, the effect is modest and confined to heavy, long-term occupational exposure — not to eating food with trace residues. Earlier AHS analyses from 2005 found the same broad picture: no strong link overall. A cautious person can reasonably conclude that career applicators have a real (if uncertain and small) reason for concern, while the evidence that dietary trace exposure causes lymphoma is weak to absent.

The Gut-Microbiome and Shikimate Hypothesis

One of the most discussed — and most overstated — ideas about glyphosate concerns the gut microbiome. The logic runs like this: humans lack the shikimate pathway, but many of our gut bacteria have it. So even though glyphosate cannot jam a human enzyme, it might disturb the balance of gut bacteria, and since the microbiome influences immunity, metabolism, and mood, that could ripple into all sorts of "modern diseases."

It is a plausible hypothesis, and it deserves honest handling in both directions. On the supportive side, laboratory work has shown that glyphosate really can perturb the gut bacteria of honey bees, some of which rely on the shikimate pathway, making them more vulnerable to infection. That is genuine experimental evidence that the mechanism is not imaginary.

But the leap from "disturbs bee gut bacteria in the lab" to "causes celiac disease, autism, obesity, and depression in humans" is enormous and not supported by good human data. The most famous paper making that broad leap presented no new experimental results — it was a hypothesis review — and is widely regarded by microbiologists as speculative. Human studies are few, small, and inconsistent, and doses that clearly shift bacteria in a dish are often far above what people ingest. The honest status: the shikimate/microbiome idea is a legitimate open scientific question that merits more research, not an established fact, and it should not be presented as one. Claims that glyphosate is "destroying the human gut" run far ahead of the evidence.

Formulations, Surfactants, and "Glyphosate Alone"

When you buy a weed-killer, you are not buying pure glyphosate. You are buying a formulation that also contains surfactants — wetting agents that help the herbicide stick to and penetrate leaves. The most studied of these, a class called POEA (polyethoxylated tallow amine), can be more toxic to cells in laboratory tests than glyphosate itself.

This matters enormously for interpreting the safety debate. Regulatory cancer classifications typically evaluate the pure active ingredient. But nobody is exposed to pure glyphosate in the field — they are exposed to the mixture. Laboratory studies have repeatedly found that complete formulations can show toxic effects at concentrations where glyphosate alone does not, which means "the active ingredient tests clean" may understate what the marketed product does. In response to exactly this concern, the European Union banned POEA surfactants from glyphosate products in 2016; other regions still permit various co-formulants.

The careful conclusion is that some of the alarm attached to "glyphosate" may really belong to the additives it is mixed with, and that studies of the active ingredient in isolation cannot fully settle the safety of the products on store shelves. This is one of the strongest and most reasonable points the concerned side makes.

Lawsuits and Settlements vs. the Science

You have almost certainly seen the headlines: juries awarding tens of millions — sometimes billions — of dollars to people with non-Hodgkin lymphoma who used Roundup, and Bayer (which bought Monsanto in 2018) setting aside more than ten billion dollars to settle tens of thousands of claims. It is tempting to read those verdicts as proof that glyphosate causes cancer. It is important to understand why that inference does not hold.

A courtroom verdict is a legal outcome, not a scientific one. Civil juries decide cases on the "preponderance of evidence" (more likely than not, essentially 51%), not the far higher bar science uses to establish causation. They weigh witnesses, sympathy, internal company documents about marketing and influence, and expert testimony from both sides — and they answer a legal question about liability, not the biological question of whether a chemical causes a disease. A company can lose lawsuits over a product that later evidence exonerates, and it can settle simply because the cost and risk of thousands of trials outweigh the cost of paying, regardless of the underlying science.

None of that means the plaintiffs' concerns are baseless, and the internal documents revealed in litigation raised uncomfortable questions about how the manufacturer engaged with regulators and science. But the settlements tell you about legal risk and corporate strategy, not about parts-per-billion on your breakfast cereal. When someone cites "$10 billion in settlements" as if it were a peer-reviewed finding, they are mixing up two different kinds of truth. The science on glyphosate is genuinely unsettled; the litigation, however dramatic, does not settle it.

Realistic Exposure and Regulatory Limits

Regulators set exposure limits with large safety margins built in. In the United States, the EPA's chronic reference dose (the amount considered safe to ingest daily over a lifetime) is 1.75 milligrams per kilogram of body weight per day; the European acceptable daily intake is lower, at 0.5 mg/kg/day. These limits are set well below the lowest dose that caused any effect in animal studies, then divided again by additional safety factors — typically 100-fold or more.

Biomonitoring studies that measure glyphosate in people's urine consistently find levels far below these limits — usually hundreds to thousands of times lower than the acceptable daily intake. In plain terms: a person eating a typical diet, even one that includes foods known to carry residues, ingests a tiny fraction of the amount regulators consider safe, which is itself a fraction of the amount that showed any harm in animals. This is the strongest quantitative argument on the reassuring side, and it is a real one.

The honest caveats: safety limits assume the science underpinning them is right, they are set for the active ingredient rather than full formulations, they generally assume no special vulnerability in fetuses or young children (an area of ongoing study), and "below the limit" is a statement about current knowledge that could shift if the underlying hazard assessment changes. So the reassurance is well-founded but conditional — it rests on the regulatory reading of the cancer question being correct.

How to Reduce Your Exposure

If you would rather lower your exposure — a completely reasonable choice given the genuine uncertainty, and one you can make without accepting either extreme — there are practical, non-alarmist steps:

The Honest Bottom Line

Here is a fair summary that respects the evidence rather than the loudest voices on either side:

Glyphosate is neither the harmless salt-water some marketing has implied nor the certain poison some campaigns claim. It is a contested chemical where careful people can weigh the same evidence and land in different places — and saying so honestly is more useful than pretending the question is closed.

Back to Table of Contents

Research Papers

  1. Guyton KZ, Loomis D, Grosse Y, et al. Carcinogenicity of tetrachlorvinphos, parathion, malathion, diazinon, and glyphosate. The Lancet Oncology. 2015;16(5):490-491. doi:10.1016/S1470-2045(15)70134-8 — The IARC Working Group summary that classified glyphosate as Group 2A, "probably carcinogenic to humans."
  2. Andreotti G, Koutros S, Hofmann JN, et al. Glyphosate Use and Cancer Incidence in the Agricultural Health Study. JNCI: Journal of the National Cancer Institute. 2018;110(5):509-516. doi:10.1093/jnci/djx233 — The largest prospective cohort of licensed applicators; found no clear association between glyphosate and non-Hodgkin lymphoma or most cancers.
  3. De Roos AJ, Blair A, Rusiecki JA, et al. Cancer incidence among glyphosate-exposed pesticide applicators in the Agricultural Health Study. Environmental Health Perspectives. 2005;113(1):49-54. doi:10.1289/ehp.7340 — An earlier AHS analysis; no strong link overall, providing historical context for the 2018 update.
  4. Zhang L, Rana I, Shaffer RM, et al. Exposure to glyphosate-based herbicides and risk for non-Hodgkin lymphoma: A meta-analysis and supporting evidence. Mutation Research/Reviews in Mutation Research. 2019;781:186-206. doi:10.1016/j.mrrev.2019.02.001 — Reported about a 41% higher NHL risk in the most highly exposed; influential but debated for its high-exposure focus.
  5. European Food Safety Authority (EFSA). Conclusion on the peer review of the pesticide risk assessment of the active substance glyphosate. EFSA Journal. 2015;13(11):4302. doi:10.2903/j.efsa.2015.4302 — The EU regulatory review concluding glyphosate is unlikely to pose a carcinogenic hazard to humans.
  6. Tarazona JV, Court-Marques D, Tiramani M, et al. Glyphosate toxicity and carcinogenicity: a review of the scientific basis of the European Union assessment and its differences with IARC. Archives of Toxicology. 2017;91(8):2723-2743. doi:10.1007/s00204-017-1962-5 — Explains why EU regulators reached a different conclusion than IARC (risk vs hazard, formulation vs active ingredient, study weighting).
  7. Portier CJ, Armstrong BK, Baguley BC, et al. Differences in the carcinogenic evaluation of glyphosate between the International Agency for Research on Cancer (IARC) and the European Food Safety Authority (EFSA). Journal of Epidemiology and Community Health. 2016;70(8):741-745. doi:10.1136/jech-2015-207005 — A critique arguing EFSA underweighted positive animal findings; represents the pro-hazard scientific view.
  8. Chang ET, Delzell E. Systematic review and meta-analysis of glyphosate exposure and risk of lymphohematopoietic cancers. Journal of Environmental Science and Health, Part B. 2016;51(6):402-434. doi:10.1080/03601234.2016.1142748 — An industry-supported meta-analysis finding no consistent association; illustrates how funding and study selection shape conclusions.
  9. Mesnage R, Defarge N, Spiroux de Vendômois J, et al. Potential toxic effects of glyphosate and its commercial formulations below regulatory limits. Food and Chemical Toxicology. 2015;84:133-153. doi:10.1016/j.fct.2015.08.012 — Argues full formulations (with surfactants) can be more toxic in cell studies than glyphosate alone.
  10. Samsel A, Seneff S. Glyphosate's Suppression of Cytochrome P450 Enzymes and Amino Acid Biosynthesis by the Gut Microbiome: Pathways to Modern Diseases. Entropy. 2013;15(4):1416-1463. doi:10.3390/e15041416 — The widely cited but heavily criticized hypothesis paper; presented no new experimental data and is regarded by many scientists as speculative.
  11. Motta EVS, Raymann K, Moran NA. Glyphosate perturbs the gut microbiota of honey bees. Proceedings of the National Academy of Sciences. 2018;115(41):10305-10310. doi:10.1073/pnas.1803880115 — Experimental evidence that glyphosate can disturb shikimate-pathway gut bacteria in bees; relevant to, but not proof of, effects on the human microbiome.
  12. Benbrook CM. Trends in glyphosate herbicide use in the United States and globally. Environmental Sciences Europe. 2016;28(1):3. doi:10.1186/s12302-016-0070-0 — Documents the roughly fifteen-fold rise in use after herbicide-tolerant crops and pre-harvest desiccation, the backdrop to modern exposure concerns.

Back to Table of Contents

Connections

Back to Table of Contents