What to check before buying
Three things, in order.
Form. Buy D3, cholecalciferol. D2, ergocalciferol, is the plant-derived
form; it raises serum 25(OH)D but research suggests it does so less
efficiently and holds it less well between doses. The only historical reason
to take D2 was that D3 came from lanolin or fish, which vegans avoid.
Lichen-derived D3 has solved that, so D2 is now a legacy product on most
shelves. If a label says “vitamin D” without specifying, put it back.
Dose per unit, not dose on the shelf. Retail vitamin D runs from 400 IU to
10,000 IU per capsule, and the high end is marketed as though more is a
feature. It is not. Most adults who need supplementation at all do well on
1,000 to 2,000 IU daily, and the useful buying rule is to pick the smallest
unit that covers your target — because you can always take two, and you can
never take half a softgel. If a clinician has set you a specific number after
a blood test, buy the unit that divides cleanly into it. Bottles at 10,000 IU
exist for correcting documented deficiency under supervision, not for
open-ended daily use.
Third-party verification. Vitamin D is a fat-soluble vitamin suspended in
oil, which means both manufacturing overage and shelf degradation are real.
Independent analyses have repeatedly found softgels delivering meaningfully
more or less than the label claims. A USP Verified, NSF, or Informed Choice
mark is the only external check available to you in a category that is
regulated as food rather than as medicine. It costs manufacturers money, which
is why cheap products that carry one are worth noticing.
A fourth thing, which is really a warning: the K2 combination is the most
common upsell in this aisle. The pitch is that vitamin K2 directs calcium
into bone and away from arteries, so D without K is incomplete or even
harmful. The mechanism is plausible and the human evidence is thin — mostly
short trials using surrogate markers such as arterial stiffness or
uncarboxylated osteocalcin, not fracture or cardiac events, and largely in
populations with low baseline K intake. It may help; it has not been shown to.
Meanwhile K2 interacts directly with warfarin and other vitamin K antagonists,
so the combination product is actively unsuitable for a group of readers who
often do not realise it. Paying a 60 percent premium for an unquantified K2
addition is not a decision the current evidence supports.
How we evaluated
We compared widely available vitamin D products on five criteria: form (D3
only), dose granularity, third-party verification, carrier oil and excipient
transparency, and cost per day at typical retail pricing. We also read
customer reports for the practical failures that do not show up on a label —
softgels leaking, capsules sticking together in humidity, bottles arriving
close to their expiry date.
We have not lab-tested these products ourselves. We do not run assays, we
do not send samples to a laboratory, and nothing here should be read as an
independent measurement of what is inside a capsule. Our evaluation is
document-based: published certifications, ingredient panels, manufacturer
disclosures, and the published independent testing programs that already
exist. Where a product’s testing claim rests only on the brand’s own internal
program, we say so in the cons.
Ratings follow the methodology on our How We Test Products page. We earn
commission on some links. Commissions never affect ranking, inclusion, or the
contents of the cons list — the cheapest product here pays us the least.
What the evidence supports — and where it stops
Correcting a genuine deficiency has been shown to matter. In older adults with
low measured 25(OH)D, vitamin D combined with calcium has been shown to reduce
fracture risk, and it remains the standard treatment for osteomalacia and for
the bone pain and muscle weakness that accompany severe deficiency. That is a
real, well-supported clinical use.
Almost everything else is weaker than the marketing suggests. Large randomised
trials in unselected adults — people who were not deficient to begin with —
did not find that supplementation reduced cancer incidence, cardiovascular
events, or total fractures. Research suggests some benefit for falls and
muscle function in deficient older adults, though the falls literature is
genuinely mixed and a few trials using very high intermittent doses found
more falls, not fewer. Vitamin D may help with seasonal mood in people who
are deficient in winter, but the trial evidence is inconsistent and the effect,
where it appears, is small. The respiratory-infection story has weakened
considerably as better-powered trials reported.
The through-line is unglamorous: the benefit of vitamin D is a function of
where you start. Moving someone from deficient to sufficient does real work.
Moving someone from sufficient to more-than-sufficient does approximately
nothing, and at sustained very high intakes it starts doing harm through
hypercalcemia. This is why the single highest-value action in this category is
a blood test, not a purchase. The full grading is in the evidence review
linked below.
FAQ
How much should I take? If you have a tested level, follow the target your
clinician sets. If you do not, 1,000–2,000 IU daily is the range most guidance
converges on for adults with limited sun exposure, and it is low enough to be
safe without monitoring. The honest answer is that the right dose is a
personal number, and you can find it out for the price of one blood test.
Does it need to be taken with food? Yes, ideally with a meal containing
some fat. Absorption of fat-soluble vitamins improves substantially with a
fat-containing meal, and it costs you nothing to move the capsule from your
morning coffee to your lunch.
Is 5,000 IU a day safe? For most healthy adults it is generally tolerated
over the medium term, but it is more than most people need and it removes your
margin for error if you are also taking a multivitamin, a fortified protein
powder, and cod liver oil without adding them up. Sustained intakes well above
this can cause hypercalcemia. If you want to be at 5,000 IU, that is a good
reason to be testing your level rather than a reason to skip testing.