EPA Position on Glyphosate
What is the EPA's current position on glyphosate?
EPA's position is that glyphosate stays registered and is "safe when used according to the label," and it still holds its 2017 finding that glyphosate is "not likely to be carcinogenic to humans." There is no registration review decision in force, though. The February 2020 Interim Decision was withdrawn in September 2022, and EPA is now updating its human health and cancer assessment, which it expects to finish in late 2026 (EPA glyphosate page, last updated August 26, 2026).
Approach: I checked EPA's pesticide chemical register and the 40 CFR Part 180 tolerances, then read EPA's official glyphosate page for the review status. My datasets don't hold the registration review decision documents themselves.
Registration status
- The parent glyphosate record is listed under Registration Review Case #0178 (PC code 417300) [1].
- The isopropylamine salt is listed as "Registered" (PC code 103601) [2].
- Food residue tolerances are in force under 40 CFR § 180.364. Compliance is measured on glyphosate alone, and the tolerances cover residues from glyphosate and its isopropylamine, ethanolamine, dimethylamine, ammonium and potassium salts. For example, yacon tuber has a tolerance of 0.2 ppm and no expiration date [3].
Human health and cancer
- EPA's 2017 assessment found no risks of concern from dietary exposure (including drinking water), residential use, aggregate exposure, or exposure of farmworkers and nearby communities. That includes infants, children and pregnant women.
- EPA says its cancer conclusion differs from IARC's "probably carcinogenic to humans" because the two differ in scope and method. It also says its conclusion matches those of EFSA, ECHA, Canada's PMRA, Australia's APVMA, Japan's Food Safety Commission and JMPR.
- Glyphosate completed Tier 1 endocrine screening with no indication it interacts with the estrogen, androgen or thyroid pathways.
- EPA says the withdrawal of the Interim Decision "does not automatically mean" its finding on cancer "is incorrect or cannot be used as support for a future decision."
- On recent literature:
- The 2000 Williams et al. review, retracted in November 2025, "has no impact" on EPA's assessment, and EPA will not rely on it.
- EPA will review the 2025 Panzacchi et al. (Ramazzini Institute) study and "may consider it" in the updated assessment.
Ecological risk and endangered species
- EPA's 2015 ecological assessment found potential risks to land and aquatic plants, mammals and birds. Risk to honey bees was low for most farm uses, but uncertain at higher rates; EPA is now reviewing new European honey bee studies.
- The final biological evaluation (November 2021) found glyphosate "likely to adversely affect" 93% of listed species and 96% of critical habitats. Formal Endangered Species Act consultation with the Fish and Wildlife Service and the National Marine Fisheries Service is ongoing.
- EPA is working to make spray-drift mitigation wording the same across all glyphosate labels.
Regulatory history and what's next
| Date | Event |
|---|---|
| Feb 2020 | Interim Decision issued: no human health risks, "unlikely to be a human carcinogen," interim drift mitigation |
| Jun 2022 | Ninth Circuit vacated the human health part and ordered EPA to reconsider ecological risks by Oct 1, 2022 |
| Sep 2022 | EPA missed that deadline and withdrew the whole Interim Decision |
| Aug 2026 | Open literature search published for comment (docket EPA-HQ-OPP-2009-0361; comments closed Sep 24, 2026) |
| Fall 2026 | Fish and Wildlife Service draft Biological Opinion due for public comment |
| Late 2026 | Updated cancer evaluation and human health risk assessment expected |
| Early 2027 | Public webinar on the updated assessment |
EPA says it has the authority to "take prompt regulatory action at any time" if new evidence shows a safety concern. No date has been set for consultation with the National Marine Fisheries Service on about 100 marine and anadromous species.
Sources:
Why do the EPA and IARC disagree about cancer?
Mostly because they answer different questions with different evidence. IARC rates hazard: whether glyphosate can cause cancer under any conditions. EPA looks at risk under expected real-world exposures. The two bodies also chose different studies and weighed them differently, so IARC rated glyphosate Group 2A, "probably carcinogenic to humans," while EPA concluded it is "not likely to be carcinogenic to humans."
Approach: The IARC Monographs list in my datasets confirms the classification, but it doesn't explain the reasoning. EPA's pesticide notices had nothing on the cancer question. The explanations below come from EPA's and IARC's own published documents.
The two positions
- IARC: Group 2A, Volume 112, evaluated in 2015 and published in 2017 [4]. It rests on "limited" evidence in humans and "sufficient" evidence in experimental animals, supported by "strong" evidence of genotoxicity (IARC Q&A, March 2016).
- EPA: "Not likely to be carcinogenic to humans" (2017). This was a weight-of-evidence call under EPA's 2005 Guidelines for Carcinogen Risk Assessment, reviewed by the FIFRA Scientific Advisory Panel in December 2016 (EPA testimony, Feb 2018).
Where they diverge
| Issue | EPA | IARC |
|---|---|---|
| What the conclusion means | Assesses "risk under expected real-world exposures" (EPA glyphosate page). It treats exposure as "a critical component of the cancer evaluation" (testimony). | "The IARC Monographs evaluation is a hazard classification." Actual cancer probability "will depend on factors such as the type and extent of exposure" (Q&A). |
| Which data count | Registrant studies plus open literature found by systematic review. That gave 14 animal cancer studies, against IARC's 10 (testimony). | Only public-domain data. Industry studies that were reported only as summary tables in online supplements were left out. IARC says many regulators "rely primarily on industry data... not available in the public domain" (Q&A). |
| Formulations vs. pure active ingredient | Many formulation studies get little weight: the exact product often wasn't identified, glyphosate wasn't measured, or the product isn't registered in the US. Effects "cannot be attributed to glyphosate" (EPA response to comments, Jan 2020). | Formulation studies are kept because real-world exposure is to formulated products. IARC found the evidence "strong" for both pure glyphosate and formulations (Q&A). |
| Genotoxicity | In vivo studies get more weight. Positive in vitro results don't change EPA's view because "the in vitro effects... do not lead to genotoxicity in vivo" (response to comments). | Rates genotoxicity evidence as "strong" (Q&A). |
| Animal tumours | Not dealt with tumour by tumour in the documents I opened. | Points to dose-related trends in mice, tumours seen "in the absence of toxicity," and rare tumour types (Q&A). |
| Non-Hodgkin lymphoma and the Agricultural Health Study (AHS) | The 2017 AHS update gives "additional strong support" for EPA's conclusion (testimony). When the 2018 AHS data are properly included, EPA's own meta-estimate falls from 1.27 (95% CI 1.01-1.59) to 1.14 (0.87-1.50), which is not statistically significant (response to comments). | The AHS follow-up was too short to "outweigh" the case-control studies. IARC's pooled analysis found a statistically significant association with NHL (Q&A). |
| International consensus | Cites agreement from Australia, Canada, Japan, New Zealand, EFSA, Germany, ECHA and JMPR (testimony). | Says its classification system is "recognized and used as a reference all around the world" (Q&A). |
Where this leaves EPA's finding: As covered in my previous answer, the Ninth Circuit vacated the human health part of EPA's 2020 Interim Decision in 2022, and EPA is now updating its cancer assessment. EPA says the withdrawal does not by itself make the 2017 cancer conclusion wrong (EPA glyphosate page).
Sources:
Is glyphosate banned in the US?
No. Glyphosate is not banned in the US. It remains a registered pesticide at the federal level. Its food tolerances are in force, and no state register I checked bars it. California, the most restrictive state, lists it as a carcinogen under Proposition 65, which means warnings are required but sales are not stopped.
Approach: I checked EPA's pesticide chemical register, the 40 CFR Part 180 tolerances, EPA's Section 24(c) state special-local-need registrations, California DPR's product register, the Prop 65 list and DTSC's Candidate Chemicals List. City and county ordinances are not in the data I can search, so I have not covered local-level restrictions.
Federal (EPA, under FIFRA and the FFDCA)
- Isopropylamine salt (the form used in most products): listed as "Registered," first registered December 1, 1982 [2].
- Glyphosate acid: in "Registration Review" under Case #0178 [1]. As covered earlier, EPA withdrew its 2020 Interim Decision. That leaves the review open; it did not cancel any registration.
- Food tolerances: 40 CFR § 180.364 holds 158 commodity tolerances for glyphosate, and none has an expiration or revocation date. One example is yacon tuber at 0.2 ppm [3].
State level
- State special-local-need registrations (FIFRA § 24(c)): several states have registered glyphosate products for local needs, and their status is still "Registered." For example, Mississippi's "Glyphosate 41%" (MS010030) [5] and California's "MON 301108 Herbicide" (CA230004, August 2023) [6].
- California DPR:
- Glyphosate products are still on the state register with product status "A." One example is product 46097, a 62.5% glyphosate isopropylamine salt concentrate for manufacturing use [7].
- Consumer Roundup lines also carry status A, for example "Roundup Super Concentrate Weed & Grass Killer" [8].
- Older registrations have status "B" and a product inactive date. For example, product 18481 has been inactive since December 3, 1986 [9].
- California Prop 65: glyphosate (CAS 1071-83-6) has been listed for cancer since July 7, 2017 [10]. A Prop 65 listing triggers warning obligations. It is not a prohibition.
- California DTSC: glyphosate is on the Candidate Chemicals List. Its hazard traits are carcinogenicity and nephrotoxicity, based on the CA MCLs, IARC 2A and Prop 65 listings [11]. Being on this list makes it eligible for Safer Consumer Products review, but the list itself does not restrict it.
What could change this: EPA's updated human health and cancer assessment and the Endangered Species Act consultations (covered in my first answer) could lead to new label restrictions or other regulatory action. For now, the registrations and tolerances above stand.