How much vitamin D do you need per day?

The RDA for vitamin D is 15 mcg (600 IU) a day for adults up to 70 and 20 mcg (800 IU) from 71.

These are the reference amounts used on US food labels and in dietary guidance, not personal targets. What you need shifts with age, pregnancy, medication and health conditions. The % Daily Value shown on every food page here is calculated against the FDA Daily Value for vitamin D, which is 20 µg a day.

What this means for you

Vitamin D is the one place where "just eat better" is bad advice, and it is worth understanding why before you change anything.

Almost no food contains it naturally. Oily fish does. Egg yolk, liver and cheese have traces. Everything else on the shelf that lists vitamin D has had it added at the factory. In the United States most of what people get comes from fortified milk, at about 3 micrograms a glass, against a daily reference of 20.

Look at what that means on a plate. A palm-sized piece of cooked salmon covers a bit over half the day. A drained can of sardines, about a third. Milk helps but does not solve it. To reach 20 micrograms from food with no fish at all, you are drinking seven glasses of milk a day. Nobody does that, and pretending otherwise is how nutrition advice loses people.

The other half of the supply is sunlight. Most people in the world do get some vitamin D from sun on skin, and that is true whatever anyone tells you about diet. The trouble is that the same ultraviolet light causes skin cancer, and skin cancer is the most preventable kind there is. Sunbathing for vitamin D trades one risk for another. No serious source recommends it.

One change that works

If you eat fish, oily fish twice a week genuinely moves the number. Tinned salmon, sardines and mackerel count, they keep in a cupboard, and they are cheap. If you do not eat fish, food is unlikely to get you there and the useful next step is a blood test rather than a better shopping list.

Established. A blood level of at least 50 nmol/L (20 ng/mL) is generally treated as adequate. Below 30 nmol/L (12 ng/mL) is deficiency, and deficiency causes real disease: soft bones in children, weak bones in adults.

Where 15 micrograms comes from, and what it does not promise

The recommended amount was worked out to cover almost everyone in a population, and it was checked recently in a way that should give anyone pause. When the same trial data were taken apart person by person instead of trial by trial, the intake needed to put 97.5 per cent of people above 50 nmol/L in winter came out at 92 micrograms a day. Six times the recommendation.

Nobody is telling you to take 92 micrograms. What that recalculation shows is that the official figure and the blood level people quote it for do not line up as neatly as the label suggests. Part of the reason is that the exchange rate underneath every one of these numbers is soft. Each extra 100 IU a day lifts blood 25(OH)D by about 1.7 nmol/L, and that figure varies enormously between studies. A soft input gives you a soft answer.

Unsettled. Doses tested head to head behave worse than the arithmetic predicts. In young women who were already deficient, 400 and 600 IU a day performed the same as each other, and even 800 IU a day left a third of them below the target after twelve weeks. So the question of what dose reaches what level does not have one answer, and this is the reason a blood test settles it and a number on a box does not.

Established. For the job these amounts were built for, the number holds. The World Health Organization had the calculation done directly for children under four, and 10 micrograms a day puts 97 per cent of them above the level where rickets appears. Preventing bone disease is what a recommended intake is for, and at that it works.

Not for everyone. Some groups run low reliably, whatever they eat. People with a BMI of 30 or more, people who have had gastric bypass surgery, people who keep their skin covered, and people with conditions that block fat absorption. If that is you, ask for the test instead of guessing.

One group is missing from that list and belongs on it. The recommended amount was worked out largely on white populations and it does not transfer. People with darker skin living far from the equator need two to four times as much vitamin D to reach the same blood level, so the figure on the label is the wrong place for them to start.

Myth. Vitamin D supplements have been sold as protection against fractures, falls, cancer and heart disease. Large trials tested all four and found no benefit. Pooling 81 randomised trials in 53,537 people puts the effect on fractures at a relative risk of 1.00 (0.93 to 1.07) and on falls at 0.97 (0.93 to 1.02), and the analysis is now precise enough to rule out even a five per cent reduction. That does not make vitamin D unimportant. It means a supplement is insurance against being deficient, not a shield against disease.

There is one group for whom that verdict flips, and it is worth knowing whether it is you or someone you look after. Vitamin D taken with calcium does cut hip fractures in people living in residential care, by about nine fractures per 1,000 people per year. The same pills in people living at home buy about one per 1,000. We took this apart properly in a separate review: do vitamin D supplements prevent fractures?

Two medicines change the arithmetic for the people taking them. Orlistat reduces absorption of vitamin D from food and supplements, so levels drift down rather than up. And statins cut cholesterol production, which is what the body builds vitamin D from, so they may reduce how much it makes. If you take either long term, this is a reason to have a level measured rather than to guess at a dose.

What a dose does to a blood level

Putting vitamin D into ordinary foods works, and there is a number for it now. Across 34 publications covering 2,398 adults and 1,532 children, fortified food at an average of 16 micrograms a day lifted blood levels by 21 nmol/L. The rise was bigger in people who started below 50 nmol/L, bigger at doses of 10 micrograms a day and up, and the curve flattened out near 21 micrograms. D3 did more than D2.

Food on its own has a ceiling that no table of fish will show you. Sixty-three adults in south-western Norway ate 750 g of salmon a week for eight weeks through the autumn. Their blood vitamin D fell anyway. It fell less than in the people eating cod and less than in the people eating no fish, and it still fell.

The instruction to take it with fat

Unsettled. Vitamin D is absorbed better from a meal that contains fat, and the familiar rule that it will not be absorbed without fat goes further than the evidence under it. A review of 46 studies found absorption happening without fat or an oily vehicle too. Which kind of fat has no answer at all yet, long chain against medium chain or monounsaturated against polyunsaturated, and the trial that raised the question said as much about its own explanation. Take it with a meal if that is easy. Do not rebuild the meal around it.

One form never reaches a supermarket shelf and is worth recognising if a doctor brings it up. Calcifediol, the half-processed form, raised blood 25(OH)D more than ordinary vitamin D3 in 12 of 17 studies covering 1,575 people. Two found the two forms identical and three came out the other way. That is a prescription decision rather than a shopping one, and what those studies measured was the blood level rather than anything that happened to the person carrying it.

Written from the evidence cards and the data on this page, then checked back against the sources. Bioma Learn has no human medical reviewer at this time and we say so rather than invent one. Information, not medical advice. How we verify.

Foods highest in vitamin D on this site

• Mackerel, salted — 20.2 µg per 1 piece (5-1/2" x 1-1/2" x 1/2") (101% DV)
• Salmon, sockeye, canned, drained solids — 17.9 µg per 3 oz (89% DV)
• Salmon, sockeye, cooked — 14.2 µg per 3 oz (71% DV)
• Trout, rainbow, farmed, cooked — 13.5 µg per 1 fillet (67% DV)
• Salmon, pink, canned, drained solids — 12 µg per 3 oz (60% DV)

All vitamin D sources ranked →

Full guide: Vitamin D: what supplements do for bones, and what large trials found they do not do