Should you take fish oil during pregnancy?

It may lower the chance of early birth, but the trials disagree, and the clearest signal is in women whose omega-3 levels start low. A 2018 Cochrane review found that among women taking omega-3, birth before 34 weeks fell from 4.6 to 2.7 percent (risk ratio 0.58, 9 trials, 5204 women), on evidence it rated high quality. The largest trial since then, ORIP, gave 900 mg of omega-3 from fish oil a day to Australian women. Early birth was 2.2 per cent with fish oil and 2.0 percent with vegetable oil, and the difference was not significant. In a US trial, the benefit showed most clearly in women who started with low DHA.

Unsettled. The trials disagree, and the reason looks like who was in them. Almost half the trials in the Cochrane review were in women at higher risk, and ORIP recruited women who mostly had enough omega-3 already. A 2021 meta-analysis of placebo-controlled trials found the drop in preterm birth disappeared once weaker trials were set aside. A fish oil capsule is not a sure way to prevent early birth, and it may matter for women whose diet is short of omega-3.

What the trials found

Preterm birth

The Cochrane review compared omega-3 supplements or foods with placebo or no omega-3. Birth before 37 weeks fell from 13.4 to 11.9 percent (risk ratio 0.89, confidence interval 0.81 to 0.97, 26 trials, 10,304 women). Birth before 34 weeks fell from 4.6 to 2.7 percent. Both results were rated high quality.

Later work pulled the other way. ORIP gave 900 mg a day of omega-3 from fish oil, from before 20 weeks to 34 weeks, across 5486 pregnancies at six Australian centers. Early preterm birth was 61 of 2734 pregnancies with fish oil and 55 of 2752 with vegetable oil, a risk ratio of 1.13 (0.79 to 1.63). A 2021 meta-analysis of placebo-controlled trials found an 11 percent drop in preterm birth that, after sensitivity analyses, shrank to a risk ratio of 0.92 (0.83 to 1.01). Its authors conclude that omega-3 does not reduce the risk.

For women who already had one preterm birth, a 2015 meta-analysis of 2 trials in 1080 women found no protection against another. Preterm birth was 34.5 percent with omega-3 and 39.8 percent without, risk ratio 0.81 (0.59 to 1.12).

Who may benefit most

The US ADORE trial enrolled 1100 women before 20 weeks and compared 200 mg of DHA a day with 1000 mg. In a per-protocol Bayesian analysis of the women who actually took the higher dose, early preterm birth fell from 3.45 to 1.2 percent. The drop was largest in women whose DHA was lowest at the start. This analysis counts women by what they took, not by the group they were assigned to, and there was no placebo, so it is a signal rather than proof.

Length of pregnancy and birth weight

Omega-3 lengthened pregnancy by 1.67 days on average in the Cochrane review (41 trials, 12,517 women, moderate quality). In a Kansas City trial of 350 women, 600 mg of DHA a day from mid-pregnancy added 2.9 days and 6.1 oz (172 g) of birth weight. In a Danish trial, 2.4 g a day of fish oil omega-3 in the third trimester added 2 days and 97 g in 699 mother and baby pairs. Those were secondary results of a trial designed to study asthma in children.

Pre-eclampsia

The Cochrane review found pre-eclampsia may possibly be reduced (risk ratio 0.84, 0.69 to 1.01, 20 trials, 8306 women), on low-quality evidence with an interval that touches no effect. A 2023 meta-analysis of 24 comparisons in 21,919 women found the same risk ratio, 0.84, with an interval of 0.74 to 0.96. It included quasi-randomized trials and trials of plant omega-3 (ALA) as well as fish oil.

The child's brain

Unsettled. The trials have not shown a benefit for the child's brain. The Cochrane review found very few differences in children's cognition, IQ, vision, language or behavior after their mothers took omega-3, mostly on low to very low-quality evidence.

Who should be careful

Not for everyone. Longer pregnancies have a cost. In the Cochrane review, pregnancy running past 42 weeks probably rose from 1.6 to 2.6 percent with omega-3 (risk ratio 1.61, 6 trials, 5141 women, moderate quality). In ORIP, more babies in the fish oil group were very large for their gestational age (risk ratio 1.30, 1.02 to 1.65), and minor stomach upsets were more common.

ORIP found no difference in serious adverse events, and no difference in interventions for pregnancies that ran past 41 weeks. If you have had a preterm birth before, the two trials in women like you did not show that omega-3 prevents another one.

Cod liver oil is a different product. The UK food table lists it at 18,000 micrograms of retinol per 3.5 oz (100 g), and the NIH advises women who are or may be pregnant not to take more than 3000 micrograms RAE of vitamin A from supplements a day, because preformed vitamin A above the upper limit can cause birth defects. We found no usable source on contaminants in fish oil supplements in pregnancy. This page does not cover them, and the absence of a source is not a finding of safety.

What expert bodies say

A 2024 clinical practice guideline advises women of childbearing age to get at least 250 mg a day of DHA plus EPA from food or supplements, and an extra 100 to 200 mg a day of DHA in pregnancy. For pregnant women with low omega-3 intake or low blood levels, it advises 600 to 1000 mg a day from about 20 weeks until about 37 weeks or birth. Low status is judged with a questionnaire about how much fish you eat, because blood tests for DHA are not yet standardized. We found no position from a government agency on fish oil supplements in pregnancy in the sources we searched.

How we searched

What would change this answer

• The full ADORE results analyzed by assigned group. The per-protocol result is promising, and it needs the standard analysis behind it.
• A trial that recruits only women with low omega-3 status and compares fish oil with placebo. That is the group the 2024 guideline targets, and it is the group the large trials mostly did not select.
• A standardized blood test for DHA in pregnancy, so that low status could be measured rather than estimated from a food questionnaire.
• An updated Cochrane review that includes ORIP and ADORE.
• Sources on contaminants in fish oil supplements in pregnancy, which this page cannot yet address.

Sources

1. ev-fopreg-01 · Meta-analysis or systematic review · Cochrane systematic review and meta-analysis of RCTs · n = 10,304
   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.
   
   Who: 19,927 women at low, mixed or high risk in 70 RCTs of omega-3 LCPUFA supplements or foods vs placebo or no omega-3
   Effect: preterm <37 wk RR 0.89 (95% CI 0.81 to 0.97; 26 RCTs, 10,304); early preterm <34 wk RR 0.58 (0.44 to 0.77; 9 RCTs, 5204); GRADE high
   Certainty: Before ORIP (2019) and ADORE (2021); almost half of the RCTs are women at elevated risk.
   Cochrane Database Syst Rev, 2018 · checked 2026-10-07 · we read the abstract
2. ev-fopreg-02 · Meta-analysis or systematic review · Cochrane systematic review and meta-analysis of RCTs · n = 12,517
   Mean gestational length was greater in women who received omega-3 LCPUFA (mean difference (MD) 1.67 days, 95% CI 0.95 to 2.39; 41 trials, 12,517 participants; moderate-quality evidence), and pre-eclampsia may possibly be reduced with omega-3 LCPUFA (RR 0.84, 95% CI 0.69 to 1.01; 20 trials, 8306 participants; low-quality evidence).
   
   Who: pregnant women in RCTs of omega-3 LCPUFA vs no omega-3
   Effect: gestational length MD +1.67 days (95% CI 0.95 to 2.39; 41 trials); pre-eclampsia RR 0.84 (0.69 to 1.01; 20 trials, 8306), not significant
   Certainty: Pregnancy length is moderate quality; preeclampsia is low, the CI crosses 1.
   Cochrane Database Syst Rev, 2018 · checked 2026-10-07 · we read the abstract
3. ev-fopreg-03 · Meta-analysis or systematic review · Cochrane systematic review and meta-analysis of RCTs · n = 5141
   Prolonged gestation > 42 weeks was probably increased from 1.6% to 2.6% in women who received omega-3 LCPUFA compared with no omega-3 (RR 1.61 95% CI 1.11 to 2.33; 5141 participants; 6 RCTs; moderate-quality evidence).
   
   Who: pregnant women in 6 RCTs reporting prolonged gestation
   Effect: prolonged gestation >42 wk RR 1.61 (95% CI 1.11 to 2.33); moderate-quality evidence
   Certainty: The flip side of prolonging pregnancy; ORIP found no further interventions for post-term pregnancy.
   Cochrane Database Syst Rev, 2018 · checked 2026-10-07 · we read the abstract
4. ev-fopreg-04 · Meta-analysis or systematic review · Cochrane systematic review and meta-analysis of RCTs · n = —
   For the child/adult outcomes, very few differences between antenatal omega-3 LCPUFA supplementation and no omega-3 were observed in cognition, IQ, vision, other neurodevelopment and growth outcomes, language and behaviour (mostly low-quality to very low-quality evidence).
   
   Who: children of women in RCTs of antenatal omega-3 LCPUFA vs no omega-3
   Effect: very few differences in cognition, IQ, vision, neurodevelopment, growth, language, behavior; mostly low to very low-quality evidence
   Certainty: The argument "fish oil for a child's brain" is not supported by RCT evidence; the 2026 MA with a positive result is preclinical (animals).
   Cochrane Database Syst Rev, 2018 · checked 2026-10-07 · we read the abstract
5. ev-fopreg-05 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 37 trials
   However, after sensitivity analyses, there were no significant differences in PTB and ePTB risk (PTB RR, 0.92; 95% CI, 0.83 to 1.01, ePTB RR, 0.82; 95% CI, 0.61 to 1.09).
   
   Who: pregnant women in 37 RCTs of omega-3 supplements vs placebo, searched to June 2020
   Effect: preterm RR 0.89 (95% CI 0.82 to 0.97), early preterm RR 0.73 (0.58 to 0.92); after sensitivity analyses RR 0.92 (0.83 to 1.01) and 0.82 (0.61 to 1.09)
   Certainty: Discrepancy with Cochrane: the authors conclude that omega-3 does not reduce the risk of preterm birth.
   Nutrients, 2021 · checked 2026-10-07 · we read the abstract
6. ev-fopreg-06 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 21,919
   RESULTS: In 24 comparisons (21,919 women) n-3 fatty acids played a protective role against the risk of preeclampsia (RR = 0.84, 95% CI 0.74-0.96 p = 0.008; I2 = 24%).
   
   Who: pregnant women in RCTs or quasi-RCTs of oral omega-3 (DHA, EPA or ALA) at least twice a week, 1990-2020
   Effect: preeclampsia RR 0.84 (95% CI 0.74 to 0.96; 24 comparisons, 21,919); preterm RR 0.86 (0.77 to 0.95; 15,510)
   Certainty: Includes quasi-RCTs and ALA; the Cochrane review gives a borderline low-quality estimate for preeclampsia.
   Arch Gynecol Obstet, 2023 · checked 2026-10-07 · we read the abstract
7. ev-fopreg-07 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 1080
   Women who received omega-3 had similar rates of PTB at <37 weeks of gestation (34.5% vs 39.8%; RR, 0.81; 95% CI, 0.59-1.12) and PTB at <34 weeks of gestation (12.0% vs 15.4%; RR, 0.62; 95% CI, 0.26-1.46) compared with control subjects.
   
   Who: asymptomatic singleton pregnancies with previous preterm birth, 2 RCTs
   Effect: preterm <37 wk 34.5% vs 39.8%, RR 0.81 (95% CI 0.59 to 1.12); <34 wk RR 0.62 (0.26 to 1.46)
   Certainty: Only two RCTs; no effect shown for the high recurrence-risk group.
   Am J Obstet Gynecol, 2015 · checked 2026-10-07 · we read the abstract
8. ev-fopreg-08 · Randomized controlled trial(s) · multicentre double-blind randomized trial · n = 5486
   Early preterm delivery occurred in the case of 61 of 2734 pregnancies (2.2%) in the n-3 group and 55 of 2752 pregnancies (2.0%) in the control group; the between-group difference was not significant (adjusted relative risk, 1.13; 95% confidence interval [CI], 0.79 to 1.63; P = 0.50).
   
   Who: women pregnant with single or multiple fetuses at six Australian centers, fish oil 900 mg n-3 LCPUFA daily from before 20 weeks to 34 weeks vs vegetable oil
   Effect: early preterm <34 wk 2.2% vs 2.0%, adjusted RR 1.13 (95% CI 0.79 to 1.63)
   Certainty: Australian women mostly with adequate omega-3 status; the largest RCT contradicts the 2018 Cochrane conclusion. PMID outside the corpus.
   Makrides M et al., N Engl J Med, 2019 · checked 2026-10-07 · we read the abstract
9. ev-fopreg-09 · Randomized controlled trial(s) · multicentre double-blind randomized trial · n = 5486
   There were no significant differences between the groups in the incidence of interventions in post-term (>41 weeks of gestation) deliveries, in adverse events, or in other pregnancy or neonatal outcomes, except that a higher percentage of infants born to women in the n-3 group than in the control group were very large for gestational age at birth (adjusted relative risk, 1.30; 95% CI, 1.02 to 1.65).
   
   Who: women in the ORIP trial, fish oil 900 mg n-3 LCPUFA daily vs vegetable oil
   Effect: very large for gestational age adjusted RR 1.30 (95% CI 1.02 to 1.65); more minor gastrointestinal disturbances; serious adverse events no different
   Certainty: Consistent with the Cochrane signal of more LGA (RR 1.15, borderline).
   Makrides M et al., N Engl J Med, 2019 · checked 2026-10-07 · we read the abstract
10. ev-fopreg-10 · Randomized controlled trial(s) · randomized Bayesian adaptive trial, per-protocol adherence analysis · n = 1100
    Using the Bayesian model, EPTB was reduced by 65%, from 3.45% to 1.2%, using both cut points.
    
    Who: 1100 women with singleton pregnancy enrolled before 20 weeks, 200 mg/day DHA plus placebo or 800 mg DHA, adherence by red cell DHA
    Effect: per-protocol Bayesian model: early preterm 3.45% to 1.2% (65% reduction) with adherence to 1000 mg; largest in lowest two quartiles of baseline DHA
    Certainty: Per-protocol analysis, not intention-to-treat; comparison of 1000 vs 200 mg, no placebo.
    Clin Nutr, 2023 · checked 2026-10-07 · we read the abstract
11. ev-fopreg-11 · Randomized controlled trial(s) · phase III double-blind randomized controlled trial · n = 350
    In comparison with placebo, DHA supplementation resulted in higher maternal and cord RBC-phospholipid-DHA (2.6%; P < 0.001), longer gestation duration (2.9 d; P = 0.041), and greater birth weight (172 g; P = 0.004), length (0.7 cm; P = 0.022), and head circumference (0.5 cm; P = 0.012).
    
    Who: 350 pregnant women in Kansas City taking DHA or placebo capsules from <20 weeks to birth; mean DHA intake 469 mg/day
    Effect: gestation +2.9 days (P = 0.041); birth weight +172 g (P = 0.004); fewer births <34 weeks (P = 0.025)
    Certainty: Small RCT; CI not given in the abstract.
    Am J Clin Nutr, 2013 · checked 2026-10-07 · we read the abstract
12. ev-fopreg-12 · Randomized controlled trial(s) · double-blind randomized controlled trial, secondary outcomes · n = 699
    n-3 LCPUFA compared with control was associated with a 2-d prolongation of pregnancy [median (IQR): 282 (275-288) d compared with 280 (273-286) d, P = 0.02], a 97-g higher birth weight (mean ± SD: 3601 ± 534 g compared with 3504 ± 528 g, P = 0.02), and an increased size for GA according to the Norwegian population-based growth curves-Skjærven (mean ± SD: 49.9 ± 28.3 percentiles compared with 44.5 ± 27.6 percentiles, P = 0.01).
    
    Who: pregnant women from the COPSAC2010 cohort, 2.4 g n-3 LCPUFA or olive oil daily from weeks 22-26 until 1 week after birth; 699 pairs analyzed
    Effect: median gestation 282 vs 280 days (P = 0.02); birth weight 3601 vs 3504 g (P = 0.02)
    Certainty: Secondary outcomes; the primary aim of the RCT was asthma in children.
    J Nutr, 2019 · checked 2026-10-07 · we read the abstract
13. ev-fopreg-13 · Position of an expert body · clinical practice guideline with formal consensus · n = —
    Women of childbearing age should obtain a supply of at least 250 mg/d of docosahexaenoic+eicosapentaenoic acid from diet or supplements, and in pregnancy an additional intake of ≥100 to 200 mg/d of docosahexaenoic acid.
    
    Who: women of childbearing age and pregnant women
    Effect: at least 250 mg/day DHA+EPA before pregnancy; +100-200 mg/day DHA in pregnancy
    Certainty: Discrepancy: ORIP and the 2021 meta-analysis see no effect in the general population; the guideline targets women with low status specifically. PMID outside the corpus.
    Cetin I et al., Am J Obstet Gynecol MFM, 2024 · checked 2026-10-07 · we read the abstract
14. ev-fopreg-14 · Position of an expert body · clinical practice guideline with formal consensus · n = —
    Thus, they should receive a supply of approximately 600 to 1000 mg/d of docosahexaenoic+eicosapentaenoic acid, or docosahexaenoic acid alone, given that this dosage showed significant reduction of preterm birth and early preterm birth in randomized controlled trials.
    
    Who: pregnant women with low DHA intake and/or low DHA blood levels
    Effect: 600-1000 mg/day DHA+EPA or DHA alone, starting by about 20 weeks, until about 37 weeks or birth
    Certainty: Status is determined by a fish intake questionnaire; a blood test for DHA is not yet standardized.
    Cetin I et al., Am J Obstet Gynecol MFM, 2024 · checked 2026-10-07 · we read the abstract
15. cdlo-r7-01 · Position of an expert body · food composition dataset · n = 1 dataset entry
    Retinol (µg) 18000 / Vitamin D (µg) 210.0 / n-3 poly /100g food (g) 24.40 / Vitamin E (mg) 20.00
    
    Who: UK composition table entry Oil, cod liver
    Effect: retinol 18000 mcg, vitamin D 210 mcg, n-3 PUFA 24.4 g, vitamin E 20 mg per 3.5 oz (100 g)
    Certainty: No USDA record, no second number; brands differ in vitamin A severalfold
    CoFID 2021 (Public Health England), code 17-488, OGL v3.0 · checked 2026-10-03 · we read the dataset
16. cdlo-r7-08 · Position of an expert body · fact sheet · n = —
    Total intakes of preformed vitamin A that exceed the UL as well as some retinoid medications used as topical therapies (such as isotretinoin, used to treat severe acne, and etretinate, a treatment for severe psoriasis) can cause congenital birth defects [1]. These birth defects can include malformations of the eye, skull, lungs, and heart [14]. Experts advise women who are or might be pregnant and those who are lactating not to take high doses (more than 3,000 mcg RAE [10,000 IU] daily) of vitamin A supplements [1].
    
    Who: women who are or might be pregnant and those who are lactating
    Effect: limit 3,000 mcg RAE (10,000 IU) daily
    Certainty: cod liver oil is a source of preformed vitamin A (CoFID: 18000 mcg/100 g)
    NIH Office of Dietary Supplements, Vitamin A Fact Sheet for Health Professionals · checked 2026-10-03 · we read the section