Guides · How many grafts

How many grafts you need,
from area and donor supply.

Graft count is the recipient area you want covered multiplied by the density you want on it, and then capped, usually hard, by what the donor area can safely give up. The published worked example puts a 189 cm² safe donor zone at a baseline of 65 follicular units per cm² at somewhere between 1,890 and 2,835 grafts in a single pass. Almost every disappointing plan comes from doing the first sum and skipping the second. Which is why the number cannot be read off a chart before somebody has measured the back of your head.

How the number is actually derived

Demand side, then supply side.

The demand sum is straightforward: measure the area to be covered in square centimetres, multiply by the follicular-unit density you are aiming for. A hairline needs more density than a crown to look right, and the front third of a hairline needs more than the two behind it, so the multiplication is done zone by zone rather than across the whole field at once.

The supply sum is the one that binds. Take the safe donor area in cm², multiply by the safe single-pass excision density (10 to 15 follicular units per cm² at a typical baseline of 65 to 75) and check that what remains is still 40 to 50 FU/cm², the point below which the donor area starts to read as thin rather than as hair. The published example runs it end to end: 189 cm² at a baseline of 65 supports 1,890 to 2,835 grafts and leaves a residual density of 50 to 55. If your demand sum comes to 4,000 and your supply sum comes to 2,400, the answer is 2,400 and the plan changes shape. It does not become 4,000 because you would prefer it to.

One more subtraction that rarely gets mentioned at consultation: the supply figure is a lifetime figure, not a per-session one, and pattern loss is progressive. Grafts spent on a crown at 32 are not available for a hairline at 45. A plan that spends the whole donor budget in one session on the area that bothers you today is a plan that assumes your hair loss has finished. It usually has not.

What we will not put on this page

The Norwood-stage table. We could not source one.

Search this question and you will find the same table repeatedly: stage II, 500–1,000 grafts; stage III, 1,000–1,500; and so on to stage VII. The numbers are consistent across dozens of clinic sites, which reads like corroboration until you look for the origin and find that none of them cites anything. We went looking for a primary source (a society guideline, a published series, anything with a method behind it) and did not find one. PLACEHOLDER: not yet checked against a named clinical source or measurement date.

So the table is not here. If we find a real source it will appear with the citation attached. In the meantime the honest version is the one above: your stage tells a surgeon roughly how much area is in play, and your donor area tells them what can be done about it, and only the second one is measured on you.

Excision limits, residual-density floor and the worked example: Keene SA, Rassman WR, Harris JA. Determining safe excision limits in FUE. Hair Transplant Forum International. 2018;28(1):1–11 (https://www.ishrs-htforum.org/content/28/1/1.2), measured 2026-07-28. The per-stage graft-count table referred to above is not sourced and is not published on this site.

FAQ

Common questions asked at consultation.

How many grafts for a Norwood 3?
We do not publish a per-stage number, because we could not find a primary source for one and the stage is the wrong input to derive it from. See the section on that refusal above. The stage describes what you have lost; the graft count is limited by what your donor area has kept.
Is more grafts always better?
No. Donor supply is finite and non-renewable, and every graft taken now is unavailable for loss that has not happened yet. The published safe single-pass excision density is roughly 10–15 follicular units per cm², with 40–50 FU/cm² left behind so the donor area still reads as hair.
Why does the clinic need to see my donor area first?
Because it is the binding constraint. Recipient area sets what you would like; donor density and safe area set what is available. A quote issued before the donor area is measured is quoting the first number and ignoring the second.