How Many Grafts Do I Need? Benchmarks by Zone, Density and Prices

What are grafts?

A graft is a natural follicular unit of the scalp, meaning a group of 1 to 4 hairs that grow out of one skin opening together and are harvested and reimplanted as a single unit. Bernstein and Rassman described this unit in 1995 as the surgical basic unit of hair transplantation. The field has counted in grafts rather than individual hairs ever since.

• Hair shaft: the visible individual hair
• Hair follicle: the single hair root in the scalp
• Graft (follicular unit): the natural group of 1 to 4 follicles that is harvested and implanted as one surgical unit
• Donor area: the DHT-resistant horseshoe rim at the back of the head and at the temples that the grafts come from

You often read that a graft consists of several follicles. More precisely: a graft is exactly one follicular unit, and that unit can contain 1, 2, 3 or 4 follicles. It is transplanted the way it grew. That is exactly what makes a natural look possible.

The right quantity is what decides the result. Place too few grafts and the scalp shows through between the new hairs. Place too many into a small area and the treated field looks denser than the rest of the hair, which reads as unnatural. Both are planning errors, not material problems.

Two terms get mixed up constantly, even though they describe different steps of the work. FUE is the harvesting technique used to release the units one by one from the donor area. DHI is the implantation technique in which they are placed directly with the Choi implanter. At Elithair the two steps belong together: DHI implantation is standard and included in the price.

1 graft, how many hairs?

A graft contains 1 to 4 hairs, about 2.2 hairs on average. That reference figure comes from the Bernstein Medical Center for Hair Restoration, which also provides the overall figures: around 110,000 scalp hairs in roughly 50,000 follicular units. On that basis, 3,000 grafts work out at around 6,600 individual hairs.

Which type goes where is what decides how natural the result looks. Single-hair grafts go exclusively into the front hairline, so the transition looks soft and slightly irregular, the way a grown hairline actually looks. The units with 2 to 4 hairs sit behind them and carry the density.

3,000 grafts work out at around 6,600 hairs. The average is enough for the conversion: 2,000 grafts are around 4,400 hairs, 3,500 grafts around 7,700 hairs, 5,000 grafts around 11,000 hairs. These are guide values, because the average varies from person to person, roughly between 1.8 and 2.3 hairs per graft.

Can grafts be split?

Grafts can be split, but that is fine detail work and not a standard route. If you mostly have three- and four-hair units but need single-hair grafts for the hairline, larger units can be separated under magnification. Every split raises the risk of transecting a follicle. With careful preparation under the microscope, experience puts the transection rate at around one per cent.

Why the count is still kept in grafts and not in hairs: the follicular unit is the surgical unit, it is harvested, stored and placed. A quote that counts hairs instead of grafts sounds like more and means the same or less. That is one of the things to watch for when you compare offers.

What determines how many grafts you need?

The number of grafts needed comes out of three figures: the size of the recipient area in cm², the target density in grafts per cm² and what the donor area can give up permanently. Area multiplied by target density gives the requirement, and the donor area decides how much of it is realistic. Everything else is fine-tuning.

At the same graft count, those finer factors still make a considerable difference:

• Hair calibre: thick hair covers noticeably better than fine hair at the same graft count
• Contrast between hair and skin colour: someone with fair skin and light hair visually needs fewer
• Wave: wavy hair covers more area than straight hair
• Donor density: how many follicular units per cm² there are in the horseshoe rim
• Pattern and stage: one connected bald area is planned differently from two receded corners
• Remaining hair: the more of your own hair is still standing, the less has to be added
• Age: in younger patients the final pattern is not yet foreseeable, so the planning is more conservative and more reserve is held back

The most important point is the one least likely to come up in a first consultation: what gets planned is not only today’s gap but also the picture in ten or twenty years. Androgenetic hair loss carries on after the treatment. That is why donor hair is deliberately held back instead of putting everything available into the first session.

Age changes the planning more than almost any other factor. In patients in their early to mid twenties the hair loss is in the middle of its development, and the later pattern is not yet foreseeable. If a lot of donor hair is spent on a low hairline at that stage, it will be missing later for the zones that have yet to thin.

In your younger years the responsible option is the more cautious one: a smaller target zone, a slightly higher hairline, a clearly planned graft reserve and often simply watching how things develop first. Why early-onset loss deserves particularly careful assessment is described in the article on early hair loss. It is not age alone that matters, but how stable the pattern is.

The difference lies in the method. Without measuring the area and the donor density, the graft count is estimated; with measurement it is calculated. You cannot see that in the quote. You will certainly see it in the result and in what is left over for a later session.

How many grafts do I need? Benchmarks by zone and Norwood stage

For temple recession the benchmarks are 2,000 to 3,500 grafts, for an entire top of the head 4,000 to 6,000 grafts. Most treatments fall between 2,800 and 3,500 grafts. For international context: the ISHRS Practice Census 2025 reports a worldwide average of 2,262 grafts per FUE treatment for the 2024 data year.

Guide values, not a diagnosis. The binding graft count is set by the medical hair analysis, because area, donor density and hair calibre are individual.

We deliberately do not list a fixed square-centimetre figure per Norwood stage. Such area values are not standardised across the field, and the numbers in the literature range roughly from 80 to 150 cm² for advanced stages. For a sense of scale: a standard credit card covers around 46 cm², so 80 cm² is just under two cards laid side by side, or roughly the palm of a hand without the fingers. What the area times target density calculation looks like in practice is shown in the worked example further down.

Temple recession and hairline (Norwood 2 to 3): 2,000 to 3,500 grafts

For temple recession and a receded hairline the benchmarks are 2,000 to 3,500 grafts. A mild recession tends to sit at 2,000 to 2,500, pronounced bald corners at 3,000 to 3,500. The number rises less with the area than with what you expect of the hairline, because single-hair grafts are placed there. Where your pattern sits is classified by the Norwood scale .

Frontal area and mid scalp (Norwood 4 to 5): 3,000 to 4,500 grafts

For the frontal area and the front of the mid scalp (Norwood 4) the benchmarks are 3,000 to 4,000 grafts; once the whorl is included (Norwood 5), 3,500 to 4,500 grafts. The transition is what matters: if only the gap is filled and the border with the remaining hair is not planned in, a visible edge appears. How a crown treatment is planned is covered in its own article.

The whorl uses up a disproportionate number of grafts. The hairs there grow out of one point in a star shape, which makes the area look smaller than it is, and the growth direction has to be rebuilt across the entire circle. Add patience to that: at the whorl the final picture only appears after 15 to 18 months, in the front area after about 12.

Large-area hair loss (Norwood 6 to 7): 5,000 grafts and up

With large-area hair loss the requirement is 5,000 to 6,000 grafts (Norwood 6), and 6,000 grafts and up once only the horseshoe rim is left (Norwood 7). Once the front zone and the crown have merged into a single area, the work is usually split across two sessions. Honesty matters more here than any number: the donor area sets the limit, and full coverage is rarely realistic. The sensible move is then to prioritise the front zone, because it frames the face.

Women and beard: 2,000 to 3,500 grafts

In women the benchmarks are 2,000 to 3,500 grafts, usually for a widening parting or a receded frontal line. Because the covering hair is often still there, a treatment without shaving is frequently possible. What is additionally checked in a hair transplant for women is whether the hair loss is diffuse and active.

For building a full beard, count on 2,000 to 3,000 grafts; for individual gaps, considerably fewer. Beard hair grows out of the skin at a very flat angle of roughly 10 to 15 degrees, while scalp hair stands much more upright. That is why a beard hair transplant is a discipline of its own and not the scalp routine transferred elsewhere.

Why Elithair does not start below 2,000 grafts

At Elithair every treatment starts at 2,000 grafts. We deliberately do not offer sessions of 1,000 to 1,500 grafts: on a visible area a number that small does not produce a closed picture, only a thin scattering that leads to a second treatment before long. Offers of 1,000 grafts are usually a price anchor rather than a treatment plan. So with any very cheap offer, do the maths on the area that number is supposed to cover.

How many grafts per cm²? Density and what actually looks realistic

The usual transplanted density is 30 to 45 grafts per cm², slightly more in places along the hairline. Healthy, non-thinned scalp carries around 65 to 85 follicular units per cm², depending on the region and the source. So the transplanted density is deliberately below the original and still looks full.

The reason for that is perception, not arithmetic. According to the density illusion described by Bernstein and Rassman, hair loss only becomes obvious to the eye once about half of the original density is gone. Put the other way round, roughly half the original density is enough for a visually closed impression. This is an established clinical principle, not a threshold measured in a lab.

More grafts per cm² are therefore not automatically better. Above roughly 50 to 60 grafts per cm², the risk grows that the recipient area will not supply the placed units with enough blood. A small clinical investigation by Beehner (ISHRS congress in Sydney 2005, only two patients) showed survival rates of 88 to 95% at 20 to 30 grafts per cm², while at 50 per cm² the figure dropped to 67% in one of the two patients. The number of cases is small, but the direction is accepted in practice: placed too tightly means poorer growth, not more hair.

That is why the recipient area is planned rather than filled by feel. In the Pre-Test System we measure the scalp, the blood supply and the hair structure before the treatment, and from that we derive the placement depth and the placement angle for each area. On that basis the graft survival rate at Elithair is approximately 98% according to our own evaluation. Without that kind of individual planning the survival rate can drop noticeably, because placement depth and angle then do not match the particular scalp and the follicles are supplied less well.

Do transplanted grafts fall out again?

The transplanted hair shafts largely fall out within the first two to six weeks; the grafts themselves are not lost in the process. The root stays in the scalp and pushes out new hair from month 3 to 4. That is a normal step in the hair cycle and not a loss of grafts, even if it feels exactly like one at the time.

That has to be distinguished from actual shock loss, meaning the temporary shedding of the surrounding native hairs. Okochi et al. evaluated 621 FUE patients on this in Aesthetic Plastic Surgery in 2024 and found it in 3.7%, around 1.6% in men and around 21% in women. This is a retrospective single-centre evaluation, so the figures are a reference point and not a population value.

That leaves expectation management, and one clear sentence helps there: a hair transplant does not restore the density of an 18-year-old. It creates a picture that looks full, and it lasts, because the transplanted hair comes from the DHT-resistant zone. The difference between that statement and a promise is exactly the difference between planning and selling.

How many grafts can be harvested at most?

Over an entire lifetime the maximum that can be harvested is of the order of 5,000 to 8,000 grafts, and only from the safe donor area at the back of the head and at the temples. So it is not the number you wish for that sets the limit, it is anatomy. The upper end applies only under ideal conditions, meaning a very dense donor area, thick hair and a favourable hair-to-skin contrast. With average conditions the capacity is considerably lower.

Safe here means genetically stable. Only the horseshoe rim is largely insensitive to DHT, the hormone that drives androgenetic hair loss. Anyone harvesting beyond it transplants hair that will itself fall out later. Unger defined this zone in 1994, based on an examination of 328 men over 65, as a narrow band roughly 7 cm high. Punia et al. put the entire permanent zone at around 203 cm² in the Indian Journal of Plastic Surgery in 2024.

It is best to think of the donor area as an account you can only withdraw from. Harvested follicles do not grow back in their old spot. All told there are an estimated 20,000 to 25,000 follicular units there, of which roughly 40 to 50% count as safely harvestable in the long run. Every session draws on that account.

Why more grafts are not better

More grafts are not better, because every harvest above the safe limit thins the donor area visibly and permanently. The technical term is overharvesting. With a short haircut the back of the head goes see-through, and that is often visually more uncomfortable than the original gap at the front. So the price of an oversized session does not come due immediately, but years later. How the donor area normally recovers is described in the article on the donor area after a hair transplant.

Are 6,000 or 7,000 grafts in one session possible? Numerically yes, but they are rarely sensible. In the literature anything above 3,000 follicular units already counts as a megasession (Gabel, Facial Plastic Surgery Clinics of North America 2013). The factual counter-argument is time outside the body: a harvested graft has no blood supply, and the widely cited rule of thumb in practice is a loss of around one per cent per hour in cooled storage. That is why many surgical protocols work with a target corridor of two to four hours between harvesting and placement.

Then there is what a very large session means physically. As the graft count rises, so do the length of the procedure, the number of extraction points in the donor area and the surface that gets punctured in the recipient area. In practice that means more hours in the treatment chair, a larger wound surface, more pronounced swelling in the forehead area during the first days and a longer healing phase.

That is exactly why very large zones are often split across two sessions instead of being squeezed into a single day. The gain is twofold: the grafts harvested last spend less time outside the body, and the scalp does not have to heal across the entire area at once.

If the donor area on the head is not enough, body hair can be considered as a supplement in individual cases, in practice only beard and chest hair. Beard hair for its thickness, chest hair because its growth in length adapts to the surrounding hair. Neither works as a replacement for a weak donor area, but both work as a supplement in certain zones.

And when the maths does not add up, the right answer is no treatment. According to our internal figures we turn down around 40% of enquiries after the hair analysis, because the donor area will not support the desired result or the hair loss is still too active. That is uncomfortable in a sales conversation and right for the result.

How many hair transplants can you have?

Two to three hair transplants are realistic over a lifetime. The limit is not set by the number of procedures but by the donor hair that remains. At least 12 months should lie between two sessions; practice cites 8 to 12 months, and the upper figure is the safe one.

A second procedure is not a sign that something went wrong the first time. Androgenetic hair loss carries on, the transplanted hair stays, your own hair does not. Someone who has the hairline treated at 30 may need the mid scalp at 40. What that looks like in practice is shown by the case of John Cena and his second hair transplant.

The 12 months have a factual reason. Before month 12 the result in the front area is not final, and at the whorl it takes 15 to 18 months. Anyone topping up before then is planning without data and risks placing grafts into an area that is going to close anyway. Fittingly, follow-up treatments come out considerably smaller than first treatments: an average of 1,637 grafts against 2,347, according to the ISHRS Practice Census 2025.

That leaves the honest closing point: plan the first treatment too big and you block the second. That is exactly why good planning holds donor hair back instead of putting the maximum possible number into the first session. The question is not how many grafts are possible today, but how many will last across twenty years.

What do grafts cost? Prices by graft count

At Elithair it is not the individual graft that costs, it is the graft tier. The flat rate runs from £2,849 for up to 2,000 grafts to £3,449 for 5,000 grafts and up, regardless of how many units are actually placed within that tier. That means the price is fixed before you travel, not after the treatment.

The DHI implantation and the patented NEO FUE are included, plus accommodation, transfers, an interpreter and aftercare. With us DHI is not a paid upgrade but the standard method for placement. That is the point where offers most often differ without it showing up in the price.

What costs extra is part of the same picture. The treatment without shaving (Zero Shave) carries a surcharge of £450. If it turns out on the day of treatment that more grafts make sense, a fixed flat amount is charged per 500 additional grafts, and only after prior agreement. Both amounts are stated before the treatment. If fewer are needed, the price drops accordingly.

Many clinics bill per graft instead. The total then depends on how many units are actually placed, and the final amount is only clear once the treatment is over. Where prices differ between countries, the reason is wage and location costs and not the number of grafts transplanted.

Health insurance generally does not contribute anything. Androgenetic hair loss counts as a cosmetic indication rather than an illness, so cover is normally not provided; what applies in your case is something your own insurer confirms. You will find the current tiers and the financing model on the pricing page.

What else influences the price of a hair transplant

Besides the graft count, the technique, the team’s experience and the location determine the price of a hair transplant. Where all three are priced very low, the savings have usually been made somewhere that does not show up on the invoice. So it pays not to compare the final figure but the items behind it.

The technique is the first item. FUE harvesting and DHI implantation take more hands and more time than older procedures, because every unit is harvested individually and placed individually. Robot-assisted systems are a further cost factor, but they play a minor role in the European market.

The second item is experience, and it acts directly on quality. A team that works well together works faster and more gently, which shortens the time a graft spends outside the body. The third item is location: clinics in high-wage cities have to cover local wage and rent costs, and in Istanbul those are lower. That is the actual reason for the price difference.

In the end the billing model matters more than the price per graft. Anyone who bills per graft and only fixes the graft count on the day of treatment shifts the price risk onto the patient: the final amount is not settled until the treatment is over. A flat price by graft tier turns that relationship around.

And because quality is something you document rather than assert: our clinic is inspected once or twice a month, unannounced, by the Turkish Ministry of Health. On top of that comes certification to ISO 9001:2015 by TÜV Thüringen, issued to the operating company Estelit Sağlık Hizmetleri A.Ş. under certificate number 151002611856. We do not audit our own quality; a public authority and an independent certification body do.

How is your graft count determined?

Your graft count is not estimated, it is calculated from three figures: the size of the recipient area in cm², the target density per cm² and the measurable density in your donor area. At Elithair that starts with the free hair analysis based on photos, with no travel and no appointment on site.

Which photos the hair analysis needs

The better the photos, the tighter the graft range. Four shots and five pieces of information are enough for a solid remote assessment. Take the pictures with dry, unstyled hair, otherwise the density looks higher than it is.

Why do different clinics quote different numbers for the same photo? Because they calculate with a different target density, draw the boundary of the treated area differently and, in the worst case, fit the number to the price quote instead of the other way around. So never ask only for the graft count, ask for the target density and the area it refers to.

Can the number still change on the day of treatment? Yes, upwards as well as downwards. Both are discussed beforehand, and billing then follows the fixed flat amount per 500 additional grafts described in the cost section. A clinic that will not answer this question in advance should make you suspicious.

If you want to know what your donor area can give and which graft range your pattern falls into: the hair analysis is free and non-binding, and it also ends in a no when no is the right answer.

How to recognise a solid graft figure

A solid graft figure always names three things together: the target zone, the intended density per cm² and the condition of the donor area. Anyone who names only a number has estimated, not calculated. With the following six questions you can test any offer, no matter who it comes from.

Five patterns should make you suspicious, and none of them depends on the provider: a figure given in hairs instead of grafts, a very high number without any assessment of the donor area, a graft count that is only fixed on the day of treatment, billing per graft with no upper limit, and commitments on density or result before any analysis has taken place.

This is how we handle it: the target zone and the target density are set before you travel, the graft tier determines the price, the hairline design is done 100% by the doctor, and every deviation is discussed beforehand. That is not a unique selling point, it is the minimum you should insist on anywhere.

Frequently asked questions about grafts

Conclusion: how many grafts do you really need?

Most treatments come in between 2,800 and 3,500 grafts: 2,000 to 3,500 for temple recession, 4,000 to 6,000 for an entire top of the head. Those are average values and not a specification. Every head brings a different area, a different hair calibre and a different donor density.

Quality and quantity of the donor area weigh equally here. Too few grafts placed and the scalp shows through; too many grafts harvested and you have a new problem at the back of the head. The right number lies between those two errors, and it comes out of measurement, not out of wishing.

The one sentence that carries this whole article: the right graft count is not the highest one, but the one that produces a full picture today and still holds up in ten years. Which one that is in your case is shown by the free hair analysis based on your photos.

Sources