How Vitamin Requirements Are Set: EAR, RDA, AI, UL and the Evidence Behind Them
How expert panels in the US, Canada, EU and UK turn research into vitamin reference values, and what the EAR, RDA, AI and UL are built on, and where the evidence is thin.
Content last updated . This page has not been medically reviewed.
The terms, and how each is built
| Term | What it means | How it is set |
|---|---|---|
| EAR (US, Canada), AR (EU), EAR (UK) | The intake estimated to meet the requirement of half of healthy people in a group. [1] The other half would not have needs met. [1] | From studies using a chosen "indicator of adequacy". [1] |
| RDA (US, Canada), PRI (EU), RNI (UK) | Enough for nearly all healthy people: 97 to 98 percent in the US and Canada [1]; at least 97 percent for the UK RNI [2]. | US and Canada: RDA = EAR + 2 standard deviations; if the spread is unknown, a 10 percent coefficient of variation is assumed, which gives 1.2 x EAR. [1] If no EAR can be set, no RDA is set. [1] |
| AI | A value used when an EAR cannot be set; based on experimental intakes or observed average intakes of healthy groups. [1] | Expressed with less certainty than an RDA; the AI might be numerically higher than an RDA would be. [1] |
| UL | The highest continuing daily intake likely to pose no risk to almost all individuals; it is not a recommended intake. [1] | From the highest intake with no observed adverse effect (or the lowest with one), reduced by an uncertainty factor. [3] |
The UK also has a Lower Reference Nutrient Intake, enough for only the small number of people with low needs. [2] EFSA defines its UL as the maximum level of total chronic daily intake not expected to pose a risk of adverse health effects. [4]
What the numbers are built on
The EU panel's 2017 summary lists the measure it used for each vitamin. Several rest on blood markers, not on disease outcomes.| Vitamin | Type of EU value | Basis, as summarized by EFSA |
|---|---|---|
| Vitamin C | AR and PRI | Balance of vitamin C losses and an adequate body pool, from studies in men, extrapolated to women [5] |
| Folate | AR and PRI | Serum and red blood cell folate, from controlled feeding and depletion-repletion studies [5] |
| Vitamin A | AR and PRI | Adequate retinol liver stores, using a factorial approach [5] |
| Vitamin D | AI only | Serum 25-hydroxy-vitamin D and musculoskeletal health outcomes; EFSA said the evidence did not allow an AR or PRI [5] |
| Vitamin B12 | AI only | Blood markers (holotranscobalamin, methylmalonic acid, homocysteine) plus observed intakes [5] |
| Vitamin E | AI only | Observed intake data [5] |
| Vitamin K | AI only | A value set by an earlier committee in 1993, for phylloquinone only [5] |
Examples of the margin: for vitamin C, EFSA estimated about 90 mg a day balances losses in men and, assuming a 10 percent coefficient of variation, set a PRI of 110 mg. [5] For folate it assumed 15 percent, giving 330 micrograms of dietary folate equivalents (DFE), partly to allow for people with a particular MTHFR genotype. [5] Different criteria and margins are one reason numbers differ between countries; see Vitamin intake references compared.
Frequently asked questions
What is the difference between an EAR and an RDA?
The EAR meets the needs of half of healthy people in a group; the RDA is set higher, to meet the needs of 97 to 98 percent. [1] In the US and Canadian method the RDA is the EAR plus two standard deviations, or 1.2 times the EAR when the spread is unknown. [1]
Why does a vitamin have an AI instead of an RDA?
Because the evidence was not strong enough to set an EAR. [1] In the EU summary, vitamins D, B12, E and K are examples with AIs. [5]
Who sets these values?
In the US and Canada, the Food and Nutrition Board of the Institute of Medicine, with Health Canada involved. [1] In the EU, EFSA's NDA Panel. [5] In the UK, government recommendations are based on advice from the Committee on Medical Aspects of Food Policy (COMA) and the Scientific Advisory Committee on Nutrition (SACN). [6]
Is the upper limit a safe amount to take?
No. A UL is the highest continuing intake likely to pose no risk to almost all individuals, not a recommended level. [1] Whether any amount is right for you is a decision for a clinician or pharmacist.
Are the numbers still being updated?
Yes, nutrient by nutrient. EFSA's upper-level overview, version 11 (August 2025), shows updates through 2023 and 2024 for several vitamins, with other re-evaluations ongoing. [4]
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
How to read the evidence labels
Labels used on this page: established nutritional role, insufficient evidence (where a panel says data were too limited).
Limits and uncertainty
- They are for healthy people. The US and Canadian panel says the values apply to the healthy population and that people with malabsorption or on dialysis may need more, with qualified medical and nutrition staff adapting the advice. [1] The UK guide likewise says its values make no allowance for needs created by disease. [2]
- A reference intake is not a personal need. The UK guide says the RNI is considerably higher than most people need, and that if one person's intake falls between the LRNI and the RNI it is not possible to say whether it is adequate, because their own requirement is unknown. [2] It also says that for most nutrients there were insufficient data to set the values with great confidence. [2]
- Children's values are often extrapolated. For the B vitamins and choline, US and Canadian values for children were extrapolated from adults where direct data were missing. [1]
- No UL does not mean safe. For some nutrients data are not enough to set a UL, which signals caution above recommended intakes, not that high intakes are harmless. [1] The EU summary lists several vitamins with "no adequate data to derive a UL". [4]
- What the UL covers differs by vitamin. For vitamin B6 and choline the US and Canadian UL is for total intake; for niacin and folate it may cover supplements and fortificants only. [1]
- Some groups have separate advice. For example, the US and Canadian panel made a separate recommendation for women who could become pregnant, to reduce neural tube defect risk. [1] See Folic acid and folate before and during pregnancy.
Who should ask a clinician
References
- Institute of Medicine. Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline, chapter 1 "Introduction to Dietary Reference Intakes". National Academies Press, 1998. Read 2026-10-23.
- Department of Health (UK). Dietary Reference Values: A Guide. HMSO, 1991. Read 2026-10-23.
- Institute of Medicine. Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline, chapter 3 (risk assessment model for Tolerable Upper Intake Levels). National Academies Press, 1998. Read 2026-10-23.
- European Food Safety Authority. Overview on Tolerable Upper Intake Levels as derived by the Scientific Committee on Food (SCF) and the EFSA NDA Panel. Version 11, August 2025. Read 2026-10-23.
- European Food Safety Authority. Dietary reference values for nutrients: summary report. EFSA supporting publication 2017:e15121, approved 4 December 2017. Read 2026-10-23.
- Public Health England. Government Dietary Recommendations: government recommendations for energy and nutrients for males and females aged 1 to 18 years and 19+ years. Published August 2016. Read 2026-10-23.