March 2015, Lyon
In March 2015 the International Agency for Research on Cancer, part of the World Health Organization, classified glyphosate as "probably carcinogenic to humans", its Group 2A. The panel cited limited evidence in humans for non-Hodgkin lymphoma, sufficient evidence of cancer in experimental animals, and strong evidence that glyphosate can cause genetic damage through oxidative stress.
The phrase "probably carcinogenic" travelled quickly. What travelled less quickly was IARC's own definition of its task: it identifies hazards, meaning whether something can cause cancer under some circumstances. It does not estimate how likely that is at real-world exposure. Red meat sits in the same Group 2A.
November 2015, Parma
Eight months later the European Food Safety Authority concluded that glyphosate was unlikely to pose a carcinogenic hazard to humans. The European Chemicals Agency reached a similar view in 2017 and again in 2022. The joint WHO and FAO Meeting on Pesticide Residues, known as JMPR, concluded in 2016 that glyphosate was unlikely to pose a carcinogenic risk through diet. The US Environmental Protection Agency took a comparable position, although a federal appeals court in 2022 ordered it to revisit parts of its analysis.
Why the readings diverged
Part of the gap is method. IARC relied mainly on published, peer-reviewed studies; regulators also weighed unpublished industry studies submitted for approval, which critics say tilts the evidence and defenders say adds data that would otherwise be ignored.
Part is emphasis. IARC gave weight to studies of formulated products and to mechanistic data; regulators focused on the pure active ingredient and on whether effects appeared at doses relevant to people. Neither approach is obviously wrong. They are answering adjacent questions.
What the large studies show
The single most cited human study is the Agricultural Health Study, which has followed more than 54,000 pesticide applicators in Iowa and North Carolina since the 1990s. Its 2018 analysis, published in the Journal of the National Cancer Institute, found no overall association between glyphosate use and non-Hodgkin lymphoma, though it noted a possible signal for one leukaemia subtype that needed further study.
A 2019 meta-analysis led by Luoping Zhang at the University of California, Berkeley, pooled several studies and reported a roughly 41 percent increased risk of non-Hodgkin lymphoma among the most heavily exposed people. Critics of that paper questioned how studies were combined and which exposure levels were chosen; its authors defended the approach as appropriate for detecting effects in high-exposure groups.
Who is actually exposed
Here the conversation usefully splits. For most people, exposure comes through trace residues in food, typically far below the limits set by regulators. For farm workers, landscapers and applicators handling concentrate for years, exposure can be orders of magnitude higher. A fair reading treats these as different situations. Reassurance about breakfast cereal does not automatically extend to someone spraying daily without gloves.
Weighing it up
The reassuring reading
- Most regulators worldwide find no likely cancer risk at expected exposures.
- The largest cohort study found no overall lymphoma link.
- Dietary residues are generally low.
The cautious reading
- IARC's hazard finding stands.
- Some analyses suggest risk among the most exposed.
- Formulations and long-term occupational use are less studied.
Our reading: for typical dietary exposure, the weight of regulatory evidence is reassuring. For heavy occupational use, the uncertainty is genuine, and sensible protective practice is the honest response. If you have a specific health concern, speak to a doctor rather than a website.