Food Contaminants
EWG Dirty Dozen list predicts consumer pesticide health risk
In plain terms: Does the Dirty Dozen list tell you which produce is actually risky?
Leans against Food Contaminants 🔬 Includes disconfirming
Part of: 🔍 pesticides in food
RefutedContestedStrong support
consensus score -0.53
Not really. It ranks by how often residue is detected, not by how much. When researchers ran the dose calculations on the list's own fruits and vegetables, every single one came in under the safety threshold, for adults and children alike.
📅 Last reviewed: 2026-07-29 ⓘ
Evidence ladder
How far up the ladder this claim has climbed. A high consensus on a low rung means "consistent so far," not "proven in people."
Top evidence so far: Population patterns (Observational)
MechanismIn-vitroAnimalObservationalRCTMeta-analysis
How the studies fall
0 support 24 contradict 0 tested null 30 mixed · 54 sources, 24 independent groups
What the evidence shows
Ranking produce by how OFTEN residue is detected is not the same as ranking it by dose. When researchers applied formal risk assessment to EWG's own 2022 list, using USDA residue measurements against EPA toxicological reference doses, EVERY pesticide-produce combination came in below the health-based guidance value for adults and children.
The evidence (54)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Jingkun L et al 2020 · Research Square | animal | mixed | low | EXPERIMENTAL ANIMAL dosing (mouse, preprint, unreviewed): chronic 'long-term, low-level' acetamiprid+tebuconazole exposure induced insulin resistance, obesity, chronic inflammation and NAFLD via gut-flora disruption, reversible by fecal tra |
| Dumitriu Gabur GD et al 2022 · Foods (Basel, Switzerland) | observational | contradicts | low | DIETARY route, Romanian wine (grape-derived, EWG Dirty-Dozen item): 6 pesticides measured at 0.05-0.75 ng/g in finished wine; hazard quotient and hazard index stayed below 1 for both women and men, and a PRIMo model found European adult win |
| Keklik M et al 2026 · study_type: observational | observational | mixed | moderate | DIETARY route, deterministic+Monte Carlo assessment of 182 sweet/jalapeno pepper samples, Turkiye (EWG 'hot peppers' category): cumulative chronic exposure stayed well below safety thresholds, but high-end acute scenarios flagged a potentia |
| Singh S et al 2022 · Biomedical chromatography : BMC | observational | contradicts | low | DIETARY route, tomato field trial (4 insecticides): theoretical maximum dietary intake stayed below both maximum permissible intake and ADI despite measurable initial residue deposits (0.227-0.593 mg/kg) shortly after application. |
| Maggioni DA et al 2018 · study_type: observational | observational | mixed | low | DIETARY route, Argentina national deterministic+stochastic acute exposure assessment (28 a.i./75 foods for children 2-5y; 9 a.i./59 foods for women 10-49y): using conservative deterministic methods with national MRLs as residue proxies (not |
| Selim MT et al 2023 · Molecules (Basel, Switzerland) | observational | mixed | low | DIETARY route, 801 Saudi vegetable samples (tomato, pepper, squash, eggplant, cucumber): estimated daily intakes stayed below ADI overall, but health risk index (HRI) reached 127.5 for fipronil in sweet pepper and 1.54/1.61 for difenoconazo |
| Andrade GCRM et al 2024 · Food additives & contaminants. Part B, Surveillance | observational | mixed | moderate | DIETARY route, 1063 Brazilian fruit/veg samples 2015-2022 (tomato, cabbage, orange): chronic exposure fell within 0.002-0.9% of ADI, but pyraclostrobin in orange showed potential ACUTE risk for adults at 112% of the acute reference dose. Ch |
| Renwick AG. 2002 · Pest management science | observational | mixed | low | Regulatory-science review (no new data): states MRLs are NOT directly linked to ADI/ARfD -- residues at/below the MRL do not necessarily approach the ADI/ARfD -- and notes 'despite this, there is public concern over such residues,' i.e. it |
| Hamey PY et al 1999 · Regulatory toxicology and pharmacology : RTP | observational | mixed | low | Methodological review (UK/1997 data): documents individual-unit pesticide residues in produce (carrots, apples, tomatoes) can run up to 25x higher than the bulked-sample average used in routine monitoring, meaning naive point-estimate/detec |
| Gabur GD et al 2026 · study_type: observational | observational | contradicts | low | DIETARY route (processed grape product): 7 pesticides tracked through winemaking of Feteasca neagra/Cabernet Sauvignon showed cumulative residue reduction of 76-90% by finished wine, keeping Hazard Index <1 throughout. Grape (Dirty-Dozen it |
| Zhao E et al 2022 · Environmental science and pollution research international | observational | contradicts | low | DIETARY route, Chinese greenhouse/open-field peppers (EWG 'hot peppers' category): both fungicides stayed under China's MRL (0.5/2 mg/kg), and the JMPR acute intake model gave 0-3% and 0-5% of ARfD respectively -- well under threshold regar |
| Li J et al 2022 · study_type: observational | observational | contradicts | moderate | DIETARY route, 245 strawberry samples (EWG Dirty-Dozen item), Beijing farmers'/supermarkets: 26.0% had >=1 residue detected, 4 samples exceeded China's MRL, yet worst-case hazard index was only 0.91% (adults) and 3.62% (children) -- far bel |
| Atuhaire A et al 2017 · Environmental health insights | observational | contradicts | low | DIETARY route, Ugandan tomatoes (45 farm+market samples): mancozeb-derived dithiocarbamate residues exceeded the Codex MRL (2 mgCS2/kg) in 47.4% of farm and 14% of market samples, and washing cut residues ~30%, but the dietary risk assessme |
| Yan Q et al 2025 · study_type: observational | observational | mixed | low | DIETARY route, 173 vegetable samples, Chongqing 2022-2023 (60.7% detection, 42.2% with >=2 residues): chronic risk and food-safety index for ALL detected pesticides were acceptable, but abamectin's ACUTE dietary intake risk was unacceptable |
| Bouktif Zarrouk M et al 2020 · study_type: observational | observational | mixed | low | DIETARY route, Tunisian Total Diet Study (170 pesticides, 64 food samples): 21 pesticides quantified above LOQ never exceeded ADI, but for 149 non-quantified pesticides an upper-bound substitution method theoretically exceeded ADI for 8 pes |
| Wang P et al 2024 · Food additives & contaminants. Part B, Surveillance | observational | mixed | low | DIETARY route, Gansu province vegetables, Monte Carlo simulation: chronic exposure to carbofuran (~49%), omethoate (~52%), disulfoton (~40%) of relevant thresholds framed as contributing to 'non-carcinogenic human risk,' acephate flagged as |
| Marques JMG et al 2021 · Revista de saude publica | observational | mixed | low | DIETARY route, Brazilian national theoretical maximum intake model (2008-09 consumption data x MRLs, not measured residues): of 283 pesticides, 68 (24%) had median theoretical intake EXCEEDING the ADI -- a deliberately conservative screenin |
| Acosta-Dacal A et al 2025 · study_type: observational | observational | mixed | moderate | DIETARY route, 570 fruit/veg samples Cape Verde 2017-2020 (fruit detection 63.9%, imported fruit most contaminated): cumulative hazard index was 0.416 for adults (below threshold) but APPROACHED 1.0 for children 6-11y once adjusted for body |
| Stephenson CL et al 2018 · Food additives & contaminants. Part A, Chemistry, analysis, control, exposure & risk assessment | observational | mixed | low | DIETARY route, glyphosate acute exposure modeling (EU/global deterministic + probabilistic methods): acute intakes were below 100% of the Acute Reference Dose for ALL foodstuffs EXCEPT wild fungi; probabilistic estimates found intake exceed |
| Ambrus Á et al 2023 · preprint | observational | mixed | low | DIETARY route, preprint (not peer-reviewed), Hungarian consumers 2017-2021, 2331 samples across apples/cherries/grapes/peaches/nectarines/peppers/strawberries -- ALL current EWG Dirty-Dozen items: single-residue hazard index exceeded 1 only |
| European Food Safety Authority (EFSA) et al 2021 · EFSA journal. European Food Safety Authority | observational | contradicts | moderate | DIETARY route, EFSA retrospective cumulative risk assessment (2016-2018 EU monitoring data, 10 country populations, 2D probabilistic model): cumulative dietary exposure to pesticides causing chronic acetylcholinesterase inhibition did NOT e |
| Mesnage R et al 2022 · study_type: observational | observational | mixed | moderate | DIETARY route, UK TwinsUK biomonitoring (65 twin pairs): pyrethroid/organophosphate residues detected in ALL urine samples, glyphosate in 53%; 34 associations found between urinary pesticide metabolites and gut/faecal metabolome, but author |
| Jiang Y et al 2022 · Journal of food science | observational | mixed | low | DIETARY route, 1690 Chinese radish samples 2016-2019: overall hazard quotient/cumulative risk index far below threshold, BUT %ARfD for triazophos exceeded 100% for adults/children/toddlers, and carbofuran/aldicarb/monocrotophos/parathion ex |
| Talat F et al 2023 · Environmental science and pollution research international | observational | mixed | low | DIETARY route, Pakistani cauliflower/brinjal (n=51): estimated ACUTE health risk for chlorpyrifos and cyhalothrin-lambda in cauliflower exceeded 200% of the JMPR acute reference dose for both sexes, while chronic risk for all measured endoc |
| Mesnage R et al 2017 · Frontiers in public health | observational | mixed | low | Theoretical/regulatory-science review (no new dietary data): argues EU/US pesticide risk assessment focuses almost exclusively on the named 'active principle,' while commercial-formulation adjuvants (illustrated via glyphosate, neonicotinoi |
| European Food Safety Authority (EFSA) et al 2026 · study_type: observational | observational | contradicts | high | DIETARY route, EFSA EU-wide 2024 surveillance (9,842 coordinated + 86,449 national + 39,433 import samples; 1.2-3.6% non-compliant): despite widespread detections, probability of exceeding health-based guidance values (HBGV) was estimated L |
| Kazemi M et al 2025 · study_type: observational | observational | mixed | moderate | DIETARY route, US fertility-clinic cohort (n=633, 21-45y) using the same USDA-PDP high- vs low-pesticide-residue produce classification EWG's list draws from: no association with FSH/AMH in the primary analysis, but a subgroup without prior |
| Lecerf JM et al 2023 · Journal of the science of food and agriculture | observational | contradicts | low | DIETARY route comparison of peeled vs unpeeled apples (EWG Dirty-Dozen item): peeled apples had fewer/lower residues, but even unpeeled apples never exceeded ADI, ranging 0.04-2.10% of ADI at 95th-percentile consumption (277 g/day). Detecti |
| Yu JW et al 2026 · study_type: observational | observational | contradicts | moderate | DIETARY route, field+model study of 9 pesticides on spinach (current EWG Dirty-Dozen item): all pesticides showed hazard quotients below 10% of threshold 14 days post-application, despite detectable residues (model R^2>0.8). Detection on a |
| Marku E et al 2026 · study_type: observational | observational | mixed | low | DIETARY route, EFSA PRIMo model on Albanian vegetable market survey (417 analytes): chronic exposure stayed below ADI, but ACUTE exposure exceeded the acute reference dose (ARfD) for several pesticide-vegetable combinations, particularly ch |
| Momtaz M et al 2024 · study_type: observational | observational | mixed | low | DIETARY route, Bangladesh cauliflower/cabbage/eggplant: chlorpyrifos detected in 44-80% of samples with 53-91% of those EXCEEDING the MRL, framed by authors as a 'potential health hazard.' Rule-8 flag: no ADI/ARfD dose calculation performed |
| Kim Y et al 2026 · bioRxiv | animal | mixed | low | EXPERIMENTAL ANIMAL dosing (mouse, oral; unreplicated single-lab bioRxiv preprint, not peer-reviewed): chronic mancozeb at the human ADI (30 ug/kg/day, 1 week) caused hippocampal astrocyte atrophy and hyperlocomotion, and even 0.5 ug/kg/day |
| Luo C et al 2025 · study_type: observational | observational | contradicts | moderate | DIETARY route, 387 pineapple samples China 2023-2025 (20/88 pesticides detected, 98.71% MRL-compliant): worst-case ADI% was <1%, %ARfD <60%, hazard index <1.5% -- far below the 100% risk threshold despite routine detection of cypermethrin/c |
| Kumar A et al 2026 · study_type: observational | observational | contradicts | moderate | DIETARY route field trial: despite measurable initial residues of carbendazim (1.16-2.11 mg/kg) and chlorantraniliprole in brinjal/tomato, probabilistic acute dietary exposure stayed well within ADI once pre-harvest intervals were followed |
| Tong J et al 2024 · study_type: observational | observational | contradicts | moderate | DIETARY route, 7028 vegetable samples, Shanghai 2018-2021: 29.2% had detectable residues, only 0.47% exceeded MRLs; target hazard quotient and hazard index were all <1 across 36 detected pesticides, including leafy vegetables (highest detec |
| Jacobs N et al 2024 · Critical reviews in toxicology | observational | contradicts | moderate | DIETARY route -- the direct test of this claim: applying formal risk assessment to EWG's 2022 Dirty Dozen list (USDA Pesticide Data Program concentrations x US EPA dietary HBGVs), estimated daily exposure was BELOW the health-based guidance |
| Larsson MO et al 2018 · study_type: observational | observational | contradicts | moderate | DIETARY route, Danish national cumulative Hazard Index across 47 commodities and 6 diets: combining ALL pesticides in the diet gave HI of only 16% (adults) and 44% (children) -- for adults, roughly the calculated cancer-risk-equivalent of o |
| Yuan X et al 2025 · study_type: observational | observational | contradicts | high | DIETARY route, Japan national probabilistic model (2013-2022 monitoring, 35,552 person-days incl. 861 children 1-6y): at the 99.9th percentile of acute exposure, boscalid intake was only 0.12% (general population) and 0.28% (children) of th |
| Amar A et al 2026 · study_type: observational | observational | mixed | low | DIETARY route meta-analysis (55 studies) of Moroccan apples (EWG Dirty-Dozen item): 70% of samples had detectable organophosphate residues (mean 0.45 mg/kg) and 60% had mycotoxins, framed by authors as a 'substantial food safety concern.' R |
| European Food Safety Authority (EFSA) et al 2025 · study_type: observational | observational | contradicts | high | DIETARY route, EFSA EU-wide 2023 surveillance (13,246 coordinated + 132,793 risk-based samples; 1.0-2.0% non-compliant): probabilistic acute and chronic dietary risk assessment found risk 'very low' for most EU subpopulation groups despite |
| Ludvickova EC et al 2026 · Food Chemistry | observational | mixed | low | DIETARY route, 58 freeze-dried fruit products, Czech retail market (strawberry = EWG Dirty-Dozen item): 111 pesticide residues quantified, 12 exceeded EU MRLs after processing-factor adjustment, 54% of strawberry samples had >=10 residues, |
| Åkesson A et al 2023 · study_type: observational | observational | contradicts | high | DIETARY route, well-powered Swedish prospective cohort (n=68,844, 15y follow-up), Dietary Pesticide Hazard Index (residue/ADI ratios summed): highest-exposure quintile showed INVERSE association with CVD mortality (HR 0.82, CI 0.75-0.90) an |
| Yan Z et al 2024 · Environmental science and pollution research international | observational | contradicts | moderate | DIETARY route, 500 Chinese fruit samples (peach, orange, grape, apple, strawberry -- 4 of 5 are EWG Dirty-Dozen items): 22 pesticides detected (0.1-1930 ug/kg), yet both ADI and ARfD for every detected pesticide stayed below 100%. Detection |
| Jankowska M et al 2024 · Chemosphere | observational | mixed | moderate | DIETARY route, 1182 Polish fruit/veg samples 2021-2023 incl. cherries (EWG Dirty-Dozen item): acute exposure generally low for fruit despite multi-residue frequency (cherries, 11 pesticides, 35.5-73.4% ARfD), but a dill sample with unapprov |
| Abtahi M et al 2025 · study_type: meta-analysis | meta-analysis | mixed | moderate | DIETARY route meta-analysis of Iranian food-pesticide monitoring: 58% of pesticide-food pairs lacked MRLs, 8% exceeded them; onion+tangerine (HQ>1) and lindane (ILCR 1.4x10^-4) showed unacceptable risk (est. 242 deaths/yr, 13,792 DALYs). Re |
| Skovgaard M et al 2017 · Environmental health insights | observational | contradicts | low | DIETARY route, Bolivian lettuce/onion/potato survey (n=30, 283 pesticides screened): cypermethrin/chlorpyrifos/difenoconazol/lambda-cyhalothrin detected in 50% of lettuce samples (20% above MRL), yet NO sample's concentration -- alone or co |
| Temkin AM et al 2025 · study_type: observational | observational | mixed | moderate | DIETARY route, US NHANES (n=1837): a produce-based pesticide load index (Dirty-Dozen-style ranking of 44 crops) was NOT associated with urinary pesticide biomarker rank overall (beta=0.02, 95%CI -0.34-0.38), turning positive only after excl |
| Li T et al 2025 · study_type: observational | observational | mixed | moderate | DIETARY route, 294,703 monitoring samples across 30 Chinese provinces 2021-2023, 52 vegetable types: a combined 'detection probability x harm extent' risk index (not raw detection alone) ranked leeks/celery highest-risk and found 33/52 type |
| Atzei A et al 2025 · study_type: observational | observational | contradicts | moderate | DIETARY route, 4-year tomato monitoring (360 samples, 22 pesticide residues, up to 12.3% multi-residue samples): EFSA PRIMo/hazard-quotient consumer risk assessment found NO potential health concern for adults or toddlers; authors explicitl |
| da Silva TM et al 2024 · PloS one | observational | mixed | low | DIETARY route, 120 menus across 6 Brazilian institutional restaurants: 99% of pesticide residues fell within ADI, but methomyl insecticide was flagged as a potential ingestion risk during lunch service specifically. Near-universal margin co |
| Horska T et al 2025 · Foods (Basel, Switzerland) | observational | contradicts | moderate | DIETARY route, 4-year field study of 6 triazole fungicides in apples (EWG Dirty-Dozen item), 2 cultivars: except tebuconazole, residues stayed under 30% of MRL by harvest, and calculated intake was below both ADI and ARfD even after 5 month |
| European Food Safety Authority (EFSA). 2024 · EFSA journal. European Food Safety Authority | observational | contradicts | moderate | DIETARY route, EFSA targeted regulatory review of dimethomorph MRLs: chronic and acute dietary risk assessment using the updated PRIMo model confirmed existing toxicological reference values are protective and current MRLs (including Codex- |
| Si W et al 2024 · Food additives & contaminants. Part B, Surveillance | observational | contradicts | moderate | DIETARY route, 704 leafy-vegetable samples, Shanghai (leafy greens = EWG Dirty-Dozen category): 54 pesticide residues found in 302/704 samples, but chronic dietary risk assessment found no long-term health risk; most pesticides ranked low-t |
| Benbrook CM et al 2020 · Environmental health : a global access science source | observational | mixed | low | DIETARY route, review/tool paper synthesizing US Pesticide Data Program (288,000 samples since 1992) and UK Food Standards Agency (72,000 samples since 1999) via a Dietary Risk Index (residue intake / ADI per serving): reports substantial r |
Disagree, or know a study we missed?
We grade by evidence, not opinions. The way to weigh in is to point us to a study we haven't cited (check the evidence table above first), or to flag a problem with one we have. Every submission is reviewed; if it holds up, the grade updates and shows in Science Changes Its Mind.
Opens a short form. You'll sign in with Google so submissions are tied to a real account — we don't display your identity, and we only accept a link we can verify (PubMed, DOI, ClinicalTrials.gov).
Educational only, not medical advice. Grades and scores reflect published evidence weighted by study design and quality; see the methodology.