Swordfish: top of the mercury tables

Swordfish is the fish that tops mercury surveys, and the firmest trial numbers nearby come from fish that are low in it. In a randomized trial in 137 pregnant women in Norway, 14 oz (400 g) of cod a week for 16 weeks left hair mercury not significantly different from their usual diet, a median of 554 against 485 ug/kg. In 232 Norwegian children aged 4 to 6, fatty fish lunches three times a week for 16 weeks raised hair mercury by 0.16 mg/kg (95% CI 0.11 to 0.21), which stayed below the EFSA, WHO and US EPA reference points. We found no trial that fed people swordfish. In 2001 the FDA advised pregnant women, nursing mothers, young children and women who may become pregnant not to eat it, along with shark, king mackerel and tilefish. Among 177 fish and lobster samples bought in New York State, swordfish had the highest average mercury, 1190 ng/g, above the EU limit of 1000 ng/g, though the paper does not say how many samples were swordfish. Food surveillance in Valencia, Spain, from 2011 to 2017 also put swordfish at the top, with levels falling about 7 percent a year, a trend that may reflect where the sampled fish came from and how big they were.

The numbers, per 3.5 oz (100 g)

As people eat it, per 3.5 oz (100 g)

How much is actually in it

• In 6 USDA Foundation Foods samples of wild-caught frozen swordfish, selenium averaged 60.7 ug per 3.5 oz (100 g) and fat 8.1 g, with fat ranging from 4.2 to 12.7 g. — USDA Foundation Foods, Swordfish, frozen, wild caught, 6 samples (n = 6) Evidence E source
  Selenium 60.69 ug (50.35-69.4), fat 8.12 g (4.2-12.67), vitamin B12 0.87 ug (0.27-1.7) per 3.5 oz (100 g)
• The UK composition table lists raw swordfish at 45 ug selenium, 4.0 ug vitamin B12 and 4.1 g fat per 3.5 oz (100 g), a B12 figure about 4.6 times the USDA mean. — CoFID 2021, Swordfish, flesh only, raw, code 16-221 Evidence E source
  Selenium 45 ug, vitamin B12 4.0 ug, fat 4.1 g, protein 18.0 g, 109 calories per 3.5 oz (100 g)

What a real portion looks like

• In the USDA survey food table one swordfish fillet weighs 4.9 oz (140 g) and one steak 210 g. — USDA FNDDS Fish, swordfish, food code 26149110, fdc_id 2706301 Evidence E source
  1 fillet 140 g, 1 steak 210 g, 1 cup 135 g, quantity not specified 140 g

What cooking and storage change

• In the UK composition table, grilled swordfish has 22.9 g protein and 57 ug selenium per 3.5 oz (100 g) versus 18.0 g and 45 ug raw, as water is lost in cooking. — CoFID 2021, Swordfish, flesh only, raw (16-221) and grilled (16-222) Evidence E source
  Protein 18.0 to 22.9 g, selenium 45 to 57 ug, fat 4.1 to 5.2 g, 109 to 139 calories per 3.5 oz (100 g)

What it does to your health

• In 62,941 Norwegian pregnancies, the top fifth of dietary mercury went with babies 34 g lighter and an adjusted odds ratio of 1.19 for small for gestational age. — pregnant women in the Norwegian Mother and Child Cohort (MoBa) with a validated FFQ (n = 62,941) Evidence C source
  Birth weight -34 g (95% CI -46 to -22) top vs bottom quintile of Hg exposure; SGA adjusted OR 1.19 (1.08 to 1.30); seafood 88% of Hg exposure
• A Cochrane review of 70 RCTs with 19,927 pregnant women found omega-3 fats, as supplements or foods such as fish, lowered preterm birth before 37 weeks, risk ratio 0.89. — women at low, mixed or high risk of poor pregnancy outcomes; omega-3 LCPUFA supplements or food vs placebo or no omega-3 (n = 19,927) Evidence A source
  Preterm birth <37 wk RR 0.89 (95% CI 0.81 to 0.97; 26 RCTs, 10,304 participants; high quality); early preterm <34 wk RR 0.58 (0.44 to 0.77)
• In an Italian cohort of 412 pregnant women, postpartum thyroiditis affected 23.9% of those who mostly ate swordfish versus 4.7% of those who mostly ate oily fish. — 412 pregnant women with stable fish habits, followed to 12 months postpartum (n = 412) Evidence C source
  Postpartum thyroiditis 22/92 (23.9%) swordfish group vs 4/85 (4.7%) oily fish group, P < 0.0001

Safety, limits and interactions

• Among 177 fish and lobster samples from New York State markets, swordfish had the highest average mercury, 1,190 ng/g. — 177 fish and lobster samples from stores and markets in New York State (n = 177 samples) Evidence C source
  Swordfish mean total Hg 1,190 ng/g, highest of all species
• Pooling three large prospective cohorts, each 1 ug/g rise in maternal hair mercury from prenatal methylmercury went with 0.7 fewer IQ points in children. — children in three major prospective cohorts of prenatal methylmercury exposure (n = 3 cohorts) Evidence C source
  -0.7 IQ points per 1 ug/g maternal hair mercury at parturition; plausible range 0 to 1.5 points
• A systematic review of 19 studies found prenatal dietary mercury consistently linked to lower birth weight and to lower scores on child developmental tests. — fetuses, neonates and children up to 8 years with measured maternal dietary methylmercury exposure (n = 19 studies) Evidence C source
  Consistent association with lower birth weight; lower neuropsychological and developmental test scores; no pooled effect size
• Among 177 fish and lobster samples bought in New York State, swordfish had the highest average mercury level, 1190 ng/g, above the EU limit of 1000 ng/g. — 177 fish and lobster samples from stores and markets in New York State (n = 177 samples) Evidence C source
  Swordfish mean Hg 1190 ng/g, 22 samples above 1000 ng/g
• In 560 fish samples from food surveillance in Valencia, Spain, 2011 to 2017, swordfish had the highest mercury of any fish group and its levels fell about 7% a year. — 560 total mercury and 206 methylmercury measurements in fish sold in Comunitat Valenciana, 2011-2017 (n = 560) Evidence C source
  Swordfish highest THg, trend -7% per year on average
• In an Italian pilot survey, mercury in swordfish reached about half of the EU maximum level, the highest share among the 12 fish and seafood products tested. — 12 fish and seafood products sold in Italy, including Atlantic bluefin tuna, hake, cod and swordfish (n = 12 products) Evidence C source
  Swordfish Hg at 50% of EU maximum level, bluefin tuna 30%, red mullet 30%

What people believe that the data does not support

• US FDA and EPA advise 8 to 12 oz a week of lower-mercury seafood in pregnancy and list swordfish among the Choices to Avoid, the fish with the highest mercury levels. — people who are pregnant or breastfeeding Evidence E source
  8 to 12 oz/week of lower-mercury seafood; Choices to Avoid (highest mercury): king mackerel, marlin, orange roughy, shark, swordfish, tilefish (Gulf of Mexico), bigeye tuna
• EFSA lists swordfish, with pike, tuna and hake, among fish species with a high mercury content. — women of childbearing age, pregnant women, toddlers and children Evidence E source
  High-mercury species: swordfish, pike, tuna, hake; women of childbearing age should meet fish benefits with low-methylmercury species and stay under the TWI
• US surveys found that awareness of mercury in fish rose from 69% in 2001 to 80% in 2006 after the FDA advised pregnant women and young children not to eat swordfish. — nationally representative US Food Safety Surveys, 4482 adults in 2001 and 2275 in 2006 (n = 6757) Evidence C source
  Awareness 69% to 80%, p < .001, women aged 18-45 less aware than other women
• In a survey of 121 pregnant women in Bermuda, 71% cut large high-mercury fish such as swordfish, but 60% also cut small low-mercury fish. — 121 pregnant women receiving prenatal care in Bermuda, 2013 (n = 121) Evidence C source
  71% reduced large high-mercury fish, 60% reduced small low-mercury fish

Who this works differently for

• In an Italian cohort of 412 pregnant women, postpartum thyroiditis affected 23.9% of those who mainly ate swordfish versus 4.7% of those who mainly ate oily fish. — pregnant women in Italy with stable fish-eating habits, followed to month 12 postpartum (n = 412) Evidence C source
  Postpartum thyroiditis 22/92 (23.9%) swordfish group vs 4/85 (4.7%) oily fish group, P < 0.0001; first-trimester TPOAb 21.7% vs 4.7%
• In 236 Italian pregnant women, those who mainly ate swordfish had the highest rates of thyroid autoantibodies through pregnancy and early postpartum, and oily fish eaters the lowest. — thyroid disease-free, nonsmoking pregnant women with stable diets; 48 swordfish, 52 oily fish, 68 mixed, 68 other fish (n = 236) Evidence C source
  TgAb and TPOAb positivity highest in swordfish group, lowest in oily fish group (P < 0.001); antibody level rose with monthly swordfish intake
• In an RCT of 137 pregnant women, 14 oz (400 g) of cod a week for 16 weeks left hair mercury not significantly above habitual diet, 554 versus 485 ug/kg. — pregnant women in Norway, gestational week 20 to 36, 400 g cod fillets per week vs habitual diet (n = 137) Evidence B source
  Post-intervention hair mercury median 554 vs 485 ug/kg (p = 0.186)
• In a Spanish birth cohort of 405 nine year olds, 13% had hair mercury above the WHO tolerable weekly intake equivalent, and swordfish was the fish most linked to higher levels. — 405 children aged 9 in the INMA birth cohort, Valencia, Spain, 2013-2014 (n = 405) Evidence C source
  Hair THg geometric mean 0.89 ug/g (0.81-0.98), 13% above PTWI equivalent
• In the same Spanish cohort at age 4, 19% of 580 children had hair mercury above the WHO tolerable weekly intake equivalent, with swordfish among the fish most tied to it. — 580 children aged 4 in the INMA birth cohort, Valencia, Spain, 2008-2009 (n = 580) Evidence C source
  Hair Hg geometric mean 1.10 ug/g (1.02-1.19), 19% above PTWI equivalent
• In NHANES 2009 to 2012 data on 5656 US youth, eating high-mercury fish such as swordfish and shark went with 2.4 times higher blood mercury per step up in eating frequency. — 5656 US children and adolescents aged 1-19 y, NHANES 2009-2012 (n = 5656) Evidence C source
  Expbeta 2.40 (1.23-4.68) for high-mercury fish frequency, fewer than 0.5% at or above 5.8 ug/L
• In a randomized trial in 137 pregnant women, eating 14 oz (400 g) of cod a week for 16 weeks, a low-mercury fish unlike swordfish, left hair mercury not significantly different from habitual diet, 554 versus 485 ug/kg. — 137 pregnant women in Norway, gestational weeks 20 to 36, 400 g cod fillets per week vs habitual diet, 16 weeks (n = 137) Evidence B source
  Hair THg median 554 (392-805) vs 485 (341-740) ug/kg, p = 0.186
• In a randomized trial in 232 Norwegian children aged 4 to 6, fatty fish lunches 3 times a week for 16 weeks raised hair mercury by 0.16 mg/kg, staying below EFSA, WHO and US EPA reference points. — 232 children aged 4-6 y in Norwegian kindergartens, fatty fish (n = 114) or meat (n = 118) lunches 3 times a week, 16 weeks (n = 232) Evidence B source
  Hair THg +0.162 (0.111-0.213) mg/kg fish vs -0.053 (-0.103 to -0.002) mg/kg meat, no association with WPPSI-III scores
• A case report describes a toddler who first developed a fish-like body odor at 10 months after eating swordfish, linked to FMO3 gene variants. — 1 boy, symptoms from 10 months of age, followed to 19 months (n = 1) Evidence D source
  Transient trimethylaminuria, resolved with gradual fish reintroduction
• A case report describes a 3 year old girl with food protein-induced enterocolitis syndrome triggered by several seafoods including swordfish, confirmed by oral food challenge. — 1 girl aged 3 in Japan (n = 1) Evidence D source
  Repeated vomiting on oral food challenge

What is still unknown

• In 917 Faroese 7 year olds, prenatal methylmercury from maternal pilot whale meat went with weaker language, attention and memory, and the links remained after excluding children whose mothers had hair mercury above 10 ug/g. — children from 1,022 consecutive singleton births in the Faroe Islands, 1986 to 1987 (n = 917) Evidence C source
  Mercury-related deficits in language, attention and memory, lesser in visuospatial and motor; persisted after excluding maternal hair mercury above 10 ug/g
• The US Dietary Guidelines seafood reviews of 44 publications found that data are insufficient to judge specific fish species by mercury content for child neurocognitive development. — 44 publications on 106237 mother-offspring pairs and 25960 children (n = 44 publications) Evidence C source
  No species-level conclusion possible

The bottom line

Pregnancy and young children are where the caution is written down. That FDA advice dates from 2001, and we could not confirm the wording of the current advice a second time. In a Spanish birth cohort, 19 percent of 580 four year olds and 13 percent of 405 nine year olds had hair mercury above the equivalent of the WHO tolerable weekly intake. Swordfish was among the fish most tied to higher levels, with diet taken from questionnaires. Among 5656 US children and teenagers, each step up in how often they ate high-mercury fish such as swordfish and shark went with 2.4 times higher blood mercury (95% CI 1.23 to 4.68), a wide interval because few of them ate these fish. In an Italian cohort of 412 pregnant women, postpartum thyroiditis affected 23.9 percent of those who mostly ate swordfish and 4.7 percent of those who mostly ate oily fish. The women chose their own fish and the finding comes from one research group, so this is an association. Two single case reports add rarer reactions. A toddler with FMO3 gene variants first developed a fishy body odor at 10 months after eating swordfish, later with any fish, and it settled as fish came back gradually. A three year old girl in Japan had food protein-induced enterocolitis syndrome, with repeated vomiting on a supervised challenge, set off by several seafoods including swordfish. Warnings about this one fish can spread to all fish. In a survey of 121 pregnant women in Bermuda, 71 percent said they had cut large high-mercury fish such as swordfish and 60 percent had also cut small low-mercury fish, and none mentioned hearing of any benefit of eating fish. US awareness of mercury in fish rose from 69 percent in 2001 to 80 percent in 2006. The reviews behind the US Dietary Guidelines, 44 publications, found the data insufficient to judge single species by mercury content for child brain development, so a verdict on swordfish in particular does not exist yet. As food, swordfish carries selenium, with numbers that differ between tables. Six USDA samples of frozen wild swordfish averaged 60.7 ug of selenium and 8.1 g of fat per 3.5 oz (100 g), and fat ran threefold, from 4.2 to 12.7 g. The UK table lists raw swordfish at 45 ug selenium, 4.1 g fat and 4.0 ug vitamin B12, about 4.6 times the USDA B12 mean of 0.87 ug, for reasons nobody has pinned down. Grilling raises protein from 18.0 to 22.9 g and selenium from 45 to 57 ug per 3.5 oz (100 g) in the UK table because water leaves the fish. We have no data on what cooking does to mercury. The US survey counts a fillet as 140 g and a steak as 210 g.

Sources

1. swf-r5-15 · reference dataset
   Swordfish, frozen, wild caught | Selenium, Se | 60.69 UG | 50.35-69.4 | 6 samples ; Total lipid (fat) | 8.123 G | 4.2-12.67 ; Vitamin B-12 | 0.87 UG | 0.27-1.7
   USDA FoodData Central, Foundation Foods, release 2026-04-30 · checked 2026-09-28
2. swf-r5-16 · reference dataset
   Swordfish, flesh only, raw | CoFID 16-221 | Selenium 45 ug | Vitamin B12 4.0 ug | Fat 4.1 g | Protein 18.0 g
   CoFID 2021, Public Health England, code 16-221 · checked 2026-09-28
3. swf-r5-18 · reference dataset
   Fish, swordfish | FNDDS code 26149110 | fdc_id 2706301 | 1 fillet 64760 = 140 g | 1 steak 62898 = 210 g | 1 cup 10205 = 135 g
   USDA FNDDS, Fish swordfish, food code 26149110, fdc_id 2706301 · checked 2026-09-28
4. swf-r5-17 · reference dataset
   Swordfish, flesh only, grilled | CoFID 16-222 | Protein 22.9 g | Selenium 57 ug | Fat 5.2 g ; Swordfish, flesh only, raw | CoFID 16-221 | Protein 18.0 g | Selenium 45 ug | Fat 4.1 g
   CoFID 2021, Public Health England, codes 16-221 and 16-222 · checked 2026-09-28
5. ev-swpr-08 · cohort
   Women in the highest quintile compared with the lowest quintile of Hg exposure delivered offspring with 34 g lower birth weight (95 % CI -46 g, -22 g) and had an increased risk of giving birth to small-for-gestational-age offspring, adjusted OR = 1·19 (95 % CI 1·08, 1·30).
   Public Health Nutr, 2014 · checked 2026-10-07
6. ev-swpr-10 · Cochrane systematic review
   Preterm birth < 37 weeks (13.4% versus 11.9%; risk ratio (RR) 0.89, 95% confidence interval (CI) 0.81 to 0.97; 26 RCTs, 10,304 participants; high-quality evidence) and early preterm birth < 34 weeks (4.6% versus 2.7%; RR 0.58, 95% CI 0.44 to 0.77; 9 RCTs, 5204 participants; high-quality evidence) were both lower in women who received omega-3 LCPUFA compared with no omega-3.
   Cochrane Database Syst Rev, 2018 · checked 2026-10-07
7. swf-r5-07 · prospective cohort
   Overall, PPT prevalence was 63/412 (15.3%), but 22/92 in group A (23.9%), 4/85 in group B (4.7%; P < 0.0001 vs. group A), 17/108 (15.7%) in group C, and 16/117 (13.7%) in group D.
   Endocrine, 2019 · checked 2026-09-28
8. ev-swpr-03 · market survey
   The highest average mercury level (1,190 ng/g) was found in swordfish.
   J Food Prot, 2010 · checked 2026-10-07
9. ev-swpr-06 · pooled cohort analysis
   It finds that prenatal MeHg exposure sufficient to increase the concentration of mercury in maternal hair at parturition by 1 microg/g decreases IQ by 0.7 points.
   Am J Prev Med, 2005 · checked 2026-10-07
10. ev-swpr-07 · systematic review
    Prenatal exposure to mercury was consistently associated with lower birth weight, but only one study reported a negative association with length at birth.
    Nutr Rev, 2022 · checked 2026-10-07
11. swf-r5-01 · market survey
    Twenty-two samples had mercury concentrations greater than the 1,000 ng/g limit set by the Commission of the European Communities. ... The highest average mercury level (1,190 ng/g) was found in swordfish.
    J Food Prot, 2010 · checked 2026-09-28
12. swf-r5-02 · surveillance data
    Swordfish, fresh tuna/albacore and canned tuna, in that order, showed the highest concentrations. ... When we analized the tendency in swordfish, we observed a 7% decrease on average per year.
    Rev Esp Salud Publica, 2023 · checked 2026-09-28
13. swf-r5-03 · contaminant survey
    Highest Hg levels were found in predatory fish. The concentrations of Hg in swordfish, Atlantic bluefin tuna and red mullet accounted for 50%, 30% and 30% of the MLs, respectively.
    Food Addit Contam Part A Chem Anal Control Expo Risk Assess, 2012 · checked 2026-09-28
14. ev-swpr-01 · agency guidance
    Those who are pregnant or breastfeeding consume between 8 and 12 ounces per week of a variety of seafood from choices that are lower in mercury.
    US FDA/EPA, Advice about Eating Fish (page updated 03/05/2024) · checked 2026-10-07
15. ev-swpr-02 · agency statement
    This particularly applies to countries where fish/seafood species with a high mercury content – such as swordfish, pike, tuna and hake – are consumed regularly.
    EFSA, Fish: scenarios indicate benefits versus risks (news, 22 January 2015) · checked 2026-10-07
16. swf-r5-08 · cross-sectional survey
    In 2001, the Food and Drug Administration (FDA) advised pregnant women, nursing mothers, young children, and women who may become pregnant not to consume shark, swordfish, king mackerel, and tilefish and recommended that these women not exceed 12 ounces of other fish per week. ... Our results indicated an increase in consumers' awareness of mercury as a problem in fish (69% in 2001 to 80% in 2006, p<.001).
    Environ Res, 2011 · checked 2026-09-28
17. swf-r5-09 · cross-sectional survey
    71% of women reported reducing consumption of large fish species with greater mercury levels, but 60% reported reduced consumption of smaller, low mercury fish. No participant mentioned hearing about the benefits of fish consumption.
    PLoS One, 2015 · checked 2026-09-28
18. ev-swpr-04 · cohort
    Overall, PPT prevalence was 63/412 (15.3%), but 22/92 in group A (23.9%), 4/85 in group B (4.7%; P < 0.0001 vs. group A), 17/108 (15.7%) in group C, and 16/117 (13.7%) in group D.
    Endocrine, 2019 · checked 2026-10-07
19. ev-swpr-05 · cohort
    Positivity rates and serum concentrations of both antibodies were the greatest in group A and the lowest in group B (P < 0.001 and P < 0.05 to < 0.001, respectively).
    Endocrine, 2016 · checked 2026-10-07
20. ev-swpr-11 · randomized controlled trial
    Post-intervention, the intervention group had median (inter-quartile range) THHg concentrations of 554 (392-805) µg/kg, and the control group 485 (341-740) µg/kg (p = 0.186).
    Environ Int, 2020 · checked 2026-10-07
21. swf-r5-04 · birth cohort
    Thirteen percent of children had THg above the equivalent to the Provisional Tolerable Weekly Intake proposed by the World Health Organization. ... Children's fish intake at 9 years-old was positively associated with THg, being swordfish, canned tuna and lean fish (i.e. hake, sea bream and sole) the most associated categories.
    Environ Int, 2019 · checked 2026-09-28
22. swf-r5-05 · birth cohort
    Nineteen percent of children had mercury concentrations above the equivalent to the Provisional Tolerable Weekly Intake proposed by WHO. ... Swordfish, lean fish and canned fish were the fish categories most associated with hair mercury concentrations.
    Environ Res, 2014 · checked 2026-09-28
23. swf-r5-06 · cross-sectional survey
    high-mercury fish (swordfish and shark) [exponentiated β coefficient (expβ): 2.40; 95% CI: 1.23, 4.68]
    J Nutr, 2015 · checked 2026-09-28
24. swf-r5-10 · randomized controlled trial
    Participants were randomly assigned to intervention (400 g of cod fillets per week) or control (continued habitual diet) groups for 16 weeks (gestational week 20-36). ... Post-intervention, the intervention group had median (inter-quartile range) THHg concentrations of 554 (392-805) µg/kg, and the control group 485 (341-740) µg/kg (p = 0.186).
    Environ Int, 2020 · checked 2026-09-28
25. swf-r5-11 · randomized controlled trial
    Children in the fish group had an increase in THHg (change 0.162, 95% CI 0.111, 0.213 mg kg-1), whereas children in the meat group had decreased THHg (-0.053, 95% CI -0.103, -0.002 mg kg-1).
    Environ Int, 2018 · checked 2026-09-28
26. swf-r5-13 · case report
    A previously healthy three-year-old boy began exhibiting a rotting fish-like body odour at 10 months of age after consuming swordfish.
    Cureus, 2025 · checked 2026-09-28
27. swf-r5-14 · case report
    This case report describes a 3-year-old Japanese girl who developed FPIES triggered by multiple seafoods, including swordfish, cod, and squid.
    Kobe J Med Sci, 2024 · checked 2026-09-28
28. ev-swpr-09 · cohort
    These associations remained after adjustment for covariates and after exclusion of children with maternal hair mercury concentrations above 10 microgram(s) (50 nmol/g).
    Neurotoxicol Teratol, 1997 · checked 2026-10-07
29. swf-r5-12 · systematic review
    No net adverse neurocognitive outcomes were reported among offspring at the highest ranges of seafood intakes despite associated increases in mercury exposures. Data are insufficient for conclusive statements regarding lactation, optimal amounts, categories or specific species characterized by mercury content and neurocognitive development
    Prostaglandins Leukot Essent Fatty Acids, 2019 · checked 2026-09-28