Critical Considerations
You are not buying a result. You are buying a number.
In America the quote is written per graft — reportedly four to ten dollars each in 2026, per a peer-reviewed US population survey indexed on PubMed Central. A rate looks like a measurement: it invites you to multiply, and multiplying feels like understanding. But the precision is in the arithmetic, not in the input. Somebody looks at your scalp, estimates how much donor hair you have, estimates how much area needs covering, and produces a figure. That figure is what you are buying, and it is also the invoice. The rate is fixed and public. The count is neither — and the count is the half that moves the total.
And the estimate has an author. The party who examines you is the party who sells you the grafts. That is not an accusation of dishonesty; it is a description of the position. Nothing in the room checks the number against anything, because there is nothing to check it against at that moment.
What the profession's own count reports
The best public count belongs to the profession itself. As of 2021 — the most recent census to publish graft counts — the ISHRS Practice Census reportedly put the average first procedure at 2,176 grafts. Not 2,500. Not 4,000. And the distribution matters more than the average: reportedly 58.1% of first procedures landed between 2,000 and 2,999 grafts, nearly 28% came in below 2,000, and 4,000 or more accounted for 1%. The same series recorded 1,842 grafts in 2010 and 2,083 in 2019 — drift, not a jump.
The census is unusually honest about itself, and that honesty belongs in any quote of it: roughly a quarter of members responded, the margin is about ±6.1 points at 95% confidence, the figures are self-reported, and in the document's own words, neither ISHRS nor Relevant Research independently verified the data respondents provided. It is the best number available. It is not a certified one.
The tier that does not match the census
Now put that distribution next to the menu. Packages abroad are commonly sold in graft tiers — 1,200, 2,500 and 4,000 — as one flat price that folds in surgery, aftercare, hotel and airport transfers, the Istanbul tier structure described in Aesthetic Surgery Journal in 2025. The tier that costs the most is selling the count that, as of 2021, reportedly one first procedure in a hundred actually used. That is the whole observation. Not fraud, not a scandal — a menu whose top item is priced against a number the profession's own census calls rare.
And the count is elastic in a way the word "graft" hides. A graft is a follicular unit: a natural cluster that may hold one hair, or two, or three. The same head of hair can honestly be described as 2,000 grafts or as 4,000, depending on how finely the tissue is divided. Divide finer and the count on the invoice rises. The donor area — the back and sides of your own head — does not grow to match it.
The number you can never audit
Petrol has a meter. Meat has a scale. Gravel leaves a pile you can measure after the fact. A graft is consumed in the act of delivery: it is counted once, during surgery, out of the patient's sight, by the party being paid — and then it is in your scalp, where it cannot be recounted. There is no standard post-operative count, and nothing in the 2022 census describes one. The only practical signal is density, and density takes roughly twelve months to appear. By then the payment has cleared and the flight home is a year old.
Why this article refuses to do the arithmetic you expect
The obvious move is to take a package price, divide it by the tier's graft count, and announce a per-graft figure — "look how cheap." We do not do that, and the reason is the point of the whole piece. That division invents a unit the seller never used. A US quote is a rate times a count, often excluding consultation, medication and follow-up. A package abroad — advertised under $3,000, all in — is not priced per graft at all; part of the spread is bundling, not discount. Line them up as if the columns match and you have stopped comparing and started assuming.
None of which means cheaper is wrong. It means the question that survives is not "which is cheaper" but "who produced this number, and how would I ever check?" One more honest figure to hold alongside it: as of 2021, ISHRS members reportedly put black-market repair work at a median of 2% of their caseload — the mean, 6%, pulled up by a few practices — and in the society's next census that mean had reportedly risen to 10% by 2024. Small overall. Moving.