EPA’s Glyphosate Assessment Is Unsafe, Unhealthy, Unlawful

This spring EPA issued its Proposed Interim Registration Review Decision, PID (Relevant documents here, and in the EPA Docket EPA-HQ-OPP-2009-0361-2340). In this document, EPA reaffirms its mis-classification of glyphosate as not likely to cause cancer—a classification that conflicts with the best available science, puts Americans in harm’s way, and fails to adhere to EPA’s policies and practices.

That’s why NRDC submitted extensive scientific and legal comments detailing to EPA why its proposal is unsafe, unhealthy, and unlawful. NRDC’s wildlife expert scientist, Dr. Sylvia Fallon and I worked together on these comments, and they cover the full range of EPA Pesticide Office failures, from impacts on Monarch butterflies and pollinators, to human health risks including cancer.

EPA Pesticide Office Cozy with Monsanto (now Bayer)

EPA’s assessment just comes right out and says that it made no changes to its risk assessment, despite thousands of public comments identifying problems (PID, p. 7). In contrast to its disregard for public input, EPA has a disturbingly close cozy relationship with Monsanto, as evidenced by internal corporate documents made public through court proceedings. Dr. Charles Benbrook submitted extensive scientific and policy comments with details and documentation of this cozy relationship between Monsanto and the EPA Office of Pesticides that is supposed to regulate Monsanto.

Cancer: Non-Hodgkin Lymphoma Risks

There are five cancer classifications in the EPA Cancer Guidelines. But, the EPA Pesticide Office continues to classify glyphosate in the 5th or last category, not likely to cause cancer. The Science Advisory Panel (SAP) disagreed, with many Panelists preferring the 3rd or middle category of “suggestive” evidence of carcinogenicity (SAP 2017, p. 47-48).  See my previous blog for details on why the EPA Pesticide Office and Monsanto are wrong, and the World Health Organization IARC cancer experts are right about the cancer risks (Sass 2017).

EPA Cancer Classifications
1. Carcinogenic to humans
2. Likely to be carcinogenic
3. Suggestive evidence of carcinogenicity (Recommended by many SAP members)
4. Inadequate information to assess
5. Not likely to be carcinogenic (EPA's selection for glyphosate)

The risk estimates reported by Zhang et al (2019) are almost identical to those reported by the World Health Organization cancer experts at IARC (2015) and in a Monsanto-sponsored meta-analysis (Change and Delzell 2016), all of which report a statistically significant increased risk of NHL (a value above 1.0) associated with real-world exposure to glyphosate-based herbicide products:Since the 2017 SAP report, three of the SAP Panelists published an updated review and meta-analysis of the epidemiologic evidence from real-world exposures to pesticide applicators and other people (Zhang et al 2019).  They reported evidence of an elevated risk of non-Hodgkin lymphoma (a blood cancer), from exposure to glyphosate-based herbicides (GBH) like Roundup.

Study author (year)

Relative Risk (RR) of NHL

95% Confidence Interval (CI)

Schinassi and Leon (2014)


1.08, 1.95

WHO IARC (2015)


1.03, 1.64

Monsanto-sponsored (2016)


1.01, 1.59

Zhang et al (2019)           


1.13, 1.75

In short, putting aside whether the risk is strong (a higher RR value) or weak (a lower RR value), there is an observed link with cancer that is statistically significant. There is no scientific justification for denying the cancer evidence.

This past spring the federal health agency, Agency for Toxic Substances and Disease Registry (ATSDR), weighed in with a robust scientific review and assessment that identified both cancer and non-cancer risks from glyphosate and GBHs. The ATSDR 2019 report clearly lays out the vast array of scientific evidence linking both pure glyphosate (rodent studies) as well as formulated GBH products (in human epidemiologic studies) like Roundup to cancer. The ATSDR is housed in the Centers for Disease Control and Prevention (CDC), and is a non-regulatory federal Agency focused on science and health. See details in my blog on the ATSDR report (Sass 2019).

Non-cancer: Developmental Risks

In addition to evidence of cancer risks, the 2019 ATSDR report identified potential developmental risks from early life exposures:

  • Developmental risks reported in studies of farm families identify a link between parental use of glyphosate and an increased risk of neural tube defects (birth defects of the brain and spine), miscarriage, preterm delivery, and small for gestational age (see ATSDR Ch2, Table 2-5, p. 40-41).
  • ATSDR summarizes a number of animal studies of early-life exposure to pure glyphosate and to GBH products with statistically significant developmental effects including testicular lesions, decreased sperm production, elevated abnormal sperm, decreased testosterone, and skeletal malformations (ATSDR, p. 14, 116).

In summary, the available evidence from the published peer-reviewed scientific literature demonstrates a potential early life susceptibility to glyphosate exposure, with possible severe and lifelong effects.

Given how important the scientific assessment of the ATSDR is to the public’s understanding of glyphosate health risks, it is no wonder that Internal Monsanto emails and documents indicate that EPA Pesticide Office staff and Monsanto appear to have coordinated an effort to stop the ATSDR report (see Sass 2019 and USRTK 2019 for details and documentation).

New Science: Autism and Endocrine Risks

There is also scientific evidence that pure glyphosate and GBH products may cause endocrine disruption, that there may be age- and sex-specific differences in the observed adverse effects, and that reproductive development may be a particularly sensitive target system.

A 2019 collaborative study from researchers in Italy, Brazil, Denmark, and in the US at the George Washington University, Mount Sinai School of Medicine reported on endocrine disruption effects in male and female mice exposed to glyphosate and GBHs from pre-birth to after puberty (Manservisi et al 2019).

A 2019 collaborative study by researchers from California medical schools used State data to estimate prenatal and infant pesticide exposure for almost 3 thousand individuals diagnosed with autism spectrum disorder (ASD) (von Ehrenstein et al 2019). They reported an elevated risk of ASD associated with prenatal pesticide exposure, with the strongest link to glyphosate (16 percent above background). The risk was even higher for ASD with intellectual disability (33 percent above background).

Neither of the above two studies are referenced in the EPA assessment; EPA  does not acknowledge either endocrine or Autism risks. However, EPA states that it will continue to review studies submitted by the registrant, Monsanto, under the Endocrine Disruptor Screening Program requirements (EPA PID, Appendix E, p. 56); personally, I think neither good science nor health-protective policies can result from a regulatory process that is so broken.

EPA Unlawfully Dismissed Child Protection Factor, FQPA

EPA is well aware that it is legally obligated to apply a default 10X margin of safety under the Food Quality Protection Act (FQPA) to protect infants and children when establishing legal pesticide residue limits in food. Such added protection is justified by the evidence of developmental harm from prenatal exposures, and by critical data gaps that ATSDR identified (see NRDC comments for details). Instead, EPA used no child protective factor at all.

If EPA had used the legally required 10X FQPA factor, most glyphosate uses would have  been determined to be unsafe—above the Level of Concern, LOC—triggering the need for regulatory action to reduce harmful exposures.





Chronic (food and water) Dietary Exposure and Risk

Below LOC, acceptable risk level

2.3X above LOC


EPA2017 HHRA, p. 21;

EPA2019 PID, p. 20-21

Short-term aggregate risk (food, water, residential exposure)

Below LOC, acceptable risk level

3.8X above LOC


EPA2017 HHRA, p. 25;

EPA2019 PID, p. 22

Short-term aggregate risk (above + dermal)

Below LOC, acceptable risk level

4.7X above LOC; UNSAFE

EPA2017Revised Cancer Ass, Appendix E

Math tricks won’t make the real risk go away, and neither will ignoring legally-required protections for infants and children.

Glyphosate in Food, Water, and People

Because glyphosate is systemic, being translocated into the plant, the herbicide and its metabolites are detected in foods, waterways and tap water, and in dust (ATSDR, p. 153). Residues in plants and foods persist, even after cooking, baking, and processing. Residues have been found in infant formula, baby foods, oat cereals, and oatmeal snack bars. Friends of the Earth and partner groups reported in 2019 that glyphosate was detected in 100% of store-purchased oat cereals tested (FOE 2019).  

Glyphosate is getting into our bodies and has been doing so at increasing levels for decades—according to research in the prestigious Journal of the American Medical Association (JAMA) (Mills et al 2017). Researchers at the University of California measured levels of glyphosate in the urine of one hundred people. The subjects had been involved in a study since the 1970’s which allowed researchers to go back and look at historical and current levels of glyphosate in urine over decades. And, the trend is rising, as glyphosate crop uses rise across the country (SASS 2017).

NRDC Recommendations to EPA

As detailed and documented in NRDC’s extensive comments, we recommend that EPA do the following:

  • EPA should impose at least a tenfold uncertainty factor as required by the FQPA. We further urge EPA to consider the extensive evidence of carcinogenicity, and to regulate it as such under the FQPA.
  • EPA should immediately grant the pending petition to prohibit preharvest uses of GBHs. 
  • EPA has ignored critical science documenting the role of glyphosate in monarch’s decline. EPA must correct its continued failure to conduct a reasonable analysis of the information available to it regarding glyphosate’s harms to monarchs and available mechanisms for reducing that harm.
  • EPA, USDA and other Agencies must work together to promote effective and affordable non-chemical weed control measures, such as integrated pest management techniques like diverse crop rotations and cover crops, improving soil health, native prairie strips, would increase agricultural milkweed habitat and help spur monarch recovery.

About the Authors

Jennifer Sass

Senior Scientist, Federal Toxics, Health and Food, Healthy People & Thriving Communities Program

Join Us

When you sign up you'll become a member of NRDC's Activist Network. We will keep you informed with the latest alerts and progress reports.