How Much Fish Is OK During Pregnancy? Guidance for Tuna, Canned Tuna, and Kinmedai, with Mercury Numbers

Short answer

Table of recommended amounts (revised June 1, 2010)

“About 80 g per serving” is roughly one serving of sashimi or one fillet (one piece of sushi or sashimi is about 15 g). The amounts in the table are for the muscle portion. The right-hand column is Urayomi’s calculation from the published means: what percentage of the weekly tolerable amount (the Food Safety Commission of Japan’s 2.0 μg/kg × 55.5 kg = 111 μg) one 80 g serving represents.

For whales and dolphins the sample counts are small, 4 to 5, so the means are unstable. One serving of bottlenose dolphin is about 4.8 times the weekly tolerable amount, so the guidance is once every 2 months.

How the recommended amounts are calculated

• Tolerable amount: the Food Safety Commission of Japan’s 2.0 μg/kg per week × 55.5 kg, the average weight of pregnant women = 111 μg per week.
• What is subtracted: after subtracting mercury from foods other than fish (0.99 μg/day) and from seafood not covered by the table (half of the 7.18 μg/day from seafood), the mercury allotted to the covered fish is 78.94 μg per week.
• Divide by each fish: divide by the fish’s mean methylmercury (for example, swordfish 78.94÷0.712 = 110.9 g/week).
• The line for inclusion in the table: fish that could not be eaten 3 times a week (240 g) at 80 g per serving were put in the table.
• In the June 2010 recalculation, the trial figures would have loosened the guidance for ecchu-baigai and yumekasago, but they were left unchanged because it is a “guideline.”
• The share changes with the yardstick. The same 80 g serving of swordfish (57.0 μg of methylmercury) is 51% of the weekly tolerable amount against the Food Safety Commission’s 2.0, and 79% against EFSA’s 1.3 (72.2 μg per week). Tolerable intakes differ by about 3 times among agencies: 3.4 (Ministry of Health and Welfare, 1973), 1.6 (JECFA, 2003), 2.0 (Food Safety Commission, 2005), and 1.3 (EFSA, 2012).

Combining fish, and eating too much (MHLW Q&A)

• If you eat two or three kinds of “once a week” fish in the same week, cut each amount to one half or one third. If you eat “once a week” and “twice a week” fish in the same week, cut each amount to one half.
• If you eat too much, cut back the following week. It is explained that one meal or one week of intake does not greatly change mercury levels in the body.
• Sample menu (Q&A): one serving of striped marlin sashimi + one serving of southern bluefin tuna sashimi (two of the “up to twice a week” fish) + fish that need no caution (salt-grilled Pacific saury, minced horse mackerel (aji no tataki), tuna salad, and so on). Pacific saury, horse mackerel, mackerel, sardine, and squid are also in the “no caution needed” group.
• If you notice the pregnancy late: the placenta generally forms in the fourth month of pregnancy. The time for mercury in the body to fall by half is about 2 months. It is said that adjusting from the point you notice can be expected to have some effect. Measuring mercury in hair is said to be unnecessary (no country abroad recommends it either).
• In the Food Safety Commission’s assessment, 99.9% of hair mercury in women aged 15 to 49 was 10 ppm or less, below the starting point for the tolerable amount (11 ppm).
• Processed foods are made from a variety of ingredients, so they are said not to need special caution. Local governments are encouraged to include the advisory in the Maternal and Child Health Handbook.

Even the same tuna is split: tuna and canned tuna, Japan and the US

The tunas covered are Pacific bluefin, southern bluefin, and bigeye. Yellowfin, albacore, and young Pacific bluefin (meji) are not covered, and canned tuna is also said to be fine in normal consumption. In products on the Japanese market, white meat is albacore and light meat is yellowfin, bigeye, and skipjack. The materials do not state which species were used for the canned tuna (oil-packed tuna flakes) in MHLW’s survey.

• Japan’s “no limit” means only that the fish is not in the advisory’s table of recommended amounts. The US concentrations are from the FDA’s published table (1990–2012), and the advice is from October 2021.
• What differs is not the concentration but how the line is drawn. Japan puts amount guidance on 16 kinds, and the US sorts fish into three categories. For bigeye, Japan says up to 80 g a week and the US says “avoid.” For yellowfin, Japan says no limit and the US says once a week. We found no document from either country explaining why they differ.
• For US canned tuna, mercury in albacore (white) (mean 0.350) is about 2.8 times that in light (0.126), and the advice is split accordingly.

Comparing with the benefits of eating fish

International assessments recognize the benefits of eating fish. Observational studies show “associations”; they cannot support a causal statement such as “eating fish raises a child’s verbal IQ.”

There is a hypothesis (in a review) that selenium protects against mercury toxicity. It is still at the stage of a hypothesis and a review, and the Food Safety Commission has also answered that it could not assess the effects of other components such as nutrients. It cannot be said that “mercury is fine if the fish contains selenium.”

Epidemiology is split, and Japanese measurements have not found an association

• Faroe Islands (1,022 births, 917 children tested at age 7): prenatal methylmercury was associated with declines in language, attention, and memory. The association was the same after excluding mothers with hair mercury above 10 μg/g.
• Seychelles (779 pairs, mothers ate fish 12 meals a week on average, mean hair mercury 6.9 ppm): did not support a neurodevelopmental risk through age 9.
• Japan (Japan Environment and Children’s Study): the median methylmercury in cord blood of 3,822 people was 7.39 ng/mL. In developmental tests at ages 2 and 4 (3,083 children), there was no significant association with mercury. The authors write that further research is needed to recommend a threshold for safe intake.
• Taiji, a town where many people eat whale and dolphin: hair mercury in 1,017 residents was higher than in 14 other areas of Japan, and 32 people (3.1%) were above 50 ppm. No one was found whose condition suggested possible poisoning, but continuing the survey is under consideration.
• For the whole diet, the hazard quotient (HQ) for methylmercury in Japanese people in 2024 was 0.24–0.36 (NIHS). 90.0% of mercury intake comes from seafood.
• Limits: concentrations by species are compiled through fiscal 2008, and total mercury and methylmercury are not from the same samples. The table of recommended amounts has not been revised since June 2010.

Sources

• MHLW, “Advisory on Seafood Consumption and Mercury for Pregnant Women”
• MHLW, text of the advisory (table of recommended amounts)
• MHLW, Q&A (revised June 2010)
• MHLW, survey results on mercury in seafood (May 2010)
• MHLW, survey on mercury and residents’ health effects in Taiji (National Institute for Minamata Disease)
• FSCJ, assessment of methylmercury (2005)
• US FDA, “Advice about Eating Fish” (October 2021)
• US FDA, “Mercury Levels in Commercial Fish and Shellfish (1990-2012)”
• FAO/WHO, “Report of the Joint FAO/WHO Expert Consultation on the Risks and Benefits of Fish Consumption” (2010)
• EU Regulation 2022/617 (maximum levels for mercury; the preamble cites EFSA’s opinion)
• Grandjean P et al. 1997 (Faroe Islands)（PMID 9392777）
• Myers GJ et al. Lancet 2003;361:1686-1692 (Seychelles)（PMID 12767734）
• Hibbeln JR et al. Lancet 2007 (ALSPAC)（PMID 17307104）
• Oken E et al. Environ Health Perspect 2005 (Project Viva)（PMID 16203250）
• Kuraoka S et al. Sci Total Environ 2024 (Japan Environment and Children’s Study)（PMID 39528210）

Last checked 2026.10.03 (against the original government documents and papers, collected and cross-checked in Gemini Notebook).

Other explainers

• Should you avoid hijiki?
• Arsenic and cadmium in brown and white rice
• Baby food, infant food, and powdered formula
• Cooking at home to reduce metals
• Differences between Japanese and overseas standards
• Fifty years of heavy metal intake in Japan
• Lead: tap water, tableware, cans, and spices
• Do detox and chelation work?