A free flap is a piece of living tissue — skin, muscle, bone, or any combination — that has been deliberately disconnected from one part of the body and reattached at another, with its own artery and vein joined under the microscope to vessels at the recipient site. Once the are open and the tissue is perfused, the flap is part of its new neighbourhood. The donor site heals separately.
The principle sounds straightforward; the technique is exacting. In a busy reconstructive unit it is also routine. Free tissue transfer is now used for breast reconstruction after mastectomy, for filling bone gaps after tumour resection in the jaw or limb, for resurfacing the lower leg after open fractures, for replacing the tongue or the floor of the mouth after head-and-neck cancer, for thumb reconstruction with a transferred toe, for , and — at the field's frontier — for transplanting hands and faces. The first successful clinical free flap was reported in 197312. The fiftieth anniversary fell quietly: by 2023 the operation had become routine practice in plastic and reconstructive surgery rather than a milestone.
This article is for clinicians outside the field who refer patients into it — general practitioners, surgical colleagues, oncologists, orthopaedic teams — and for educated patients who want to understand what they are being offered. It is not a technical manual; it is an overview of the principle, where it came from, what it is used for, and how reliable it has become.

What a free flap actually is

Reconstructive surgery proceeds by a logic that most surgeons learn as the . A wound is closed by the simplest method that will work. Direct closure first, then a , then a local flap rotated from adjacent tissue, then a regional flap that travels further on a , then — when nothing nearer will do — tissue brought from elsewhere on its own blood supply. In 1994 Gottlieb and Krieger proposed the reconstructive elevator as a corrective: choose the rung that gives the best functional and aesthetic result, not necessarily the lowest one that closes the wound3. Free flaps sit at the top of the ladder; they are not always the answer, but for some defects they are the only one.
The defining feature of a free flap is that the tissue's blood supply is severed entirely and re-established by microvascular anastomosis — the joining of an artery and at least one vein, typically with sutures finer than a human hair, under an operating microscope. This distinguishes a free flap from a , which keeps its blood supply attached, and from a , in which the patient's own amputated part is reattached to its original location.
A useful way to think about it: a free flap is a transplant of the patient's own tissue from one place to another.

A brief history

The conceptual ingredients existed long before the operation. Alexis Carrel's vascular suturing technique, for which he received the 1912 Nobel Prize in Physiology or Medicine, made small-vessel anastomosis reproducible4. The operating microscope was introduced into vascular work by Jacobson and Suarez in 19605, opening surgery on vessels too small to handle by loupe alone. Harry Buncke's experimental rabbit-ear replantation in 1966 demonstrated that microvascular anastomosis at sub-millimetre scale was clinically feasible6.
The clinical firsts followed quickly. Komatsu and Tamai in Japan reported the first successful digital replantation in 1968 — a thumb cleanly amputated and reattached with restored circulation7. Cobbett in the United Kingdom reported in 1969 the first free composite tissue transfer in a human — a great toe transferred to replace an amputated thumb8. In 1973 Daniel and Taylor in Australia transferred a flap of skin and subcutaneous tissue from the groin to a recipient site by joining the flap's artery and vein to recipient vessels — the first free flap as the term is now used12. Harii, Ohmori and Ohmori published a series of ten cases the following year, demonstrating that the technique was reproducible9.
What followed was three decades of expanding what a flap could be made of. McGregor and Morgan distinguished random-pattern from axial-pattern flaps in 197310. Mathes and Nahai classified muscle flaps by their pattern of vascular pedicles in 1981 — a five-type system that remains the standard11. Pontén the same year showed that the deep fascia carried its own usable blood supply, giving rise to the 12. Taylor and Palmer mapped the body's territories of arterial supply in 1987 and called them 13; the angiosome concept is the single most cited anatomical framework in flap surgery. Koshima and Soeda described an inferior epigastric artery skin flap raised without taking the rectus abdominis muscle in 198914 — the conceptual birth of the , in which only the small skin-supplying vessel that pierces the muscle is dissected, and the muscle itself is preserved.
By the late 1990s, perforator flaps had transformed reconstruction. By the mid-2000s, — anastomosis of vessels 0.8 mm or smaller — had opened new options for lymphatic surgery and for tissue transfer at the millimetre scale15.

The vascular logic

The reason free flaps work, and the reason some flaps work better than others, comes down to vascular anatomy. The Mathes–Nahai classification organises muscles by how many dominant pedicles supply them and how those pedicles are distributed11 — a Type V muscle (one dominant pedicle plus segmental pedicles, e.g. latissimus dorsi) is a different proposition from a Type IV (segmental pedicles only, e.g. tibialis anterior, which is unusable as a free flap). The Cormack–Lamberty classification did the same for fasciocutaneous flaps in 198416.
Taylor and Palmer's angiosome concept reframed the problem at the level of the whole body13. Each named source artery supplies a three-dimensional territory of skin, fascia, muscle and bone; adjacent angiosomes communicate through choke vessels, which can dilate to extend a flap's safe territory. The implication is that a flap's reliable size is anatomically defined — and predictable, if the surgeon knows the territory.
The perforator flap built on this. A perforator is a small vessel that pierces deep fascia to supply skin. Once it became technically possible to dissect a perforator back to its source, surgeons could raise skin and subcutaneous tissue without sacrificing the underlying muscle. This produced the modern workhorse of free tissue transfer: the for breast reconstruction, which provides the same skin and fat as the older TRAM flap without the abdominal-wall morbidity of removing the rectus muscle. The Gent consensus of 2003 standardised perforator-flap nomenclature internationally17.

Workhorse flaps

A reconstructive surgeon can name dozens of free flaps; in practice, a small group accounts for most of the work.
The , originally described by Song and colleagues in Beijing in 198418 and popularised in the West by Fu-Chan Wei's group from Chang Gung Memorial Hospital19, supplies a large area of skin and fascia from the lateral thigh on the descending branch of the lateral circumflex femoral artery. It is the field's most adaptable soft-tissue flap — used for head and neck, limb, and trunk defects — partly because the donor leaves a hidden scar and tolerates significant tissue loss without functional consequence.
The DIEP flap, first reported clinically by Allen and Treece in 199420, is the autologous standard for breast reconstruction. It uses the patient's own lower abdominal skin and fat — tissue many women have available and would prefer not to keep — to rebuild a breast that ages with the body and tolerates radiotherapy. The donor closure resembles a cosmetic abdominoplasty.
The , first described by Taylor in 1975 as a vascularised bone graft21 and adapted by Hidalgo for mandibular reconstruction in 198922, supplies up to 25 centimetres of vascularised bone with a reliable skin paddle. It is the standard reconstruction for segmental jaw defects after tumour resection, and is also used for limb-length reconstruction in long-bone defects. The lower leg accommodates the loss of the fibula well.
The , developed in China in the late 1970s and described in English by Yang Guofan in 198123 and by Soutar in 198324, supplies thin, pliable, often hairless skin on a long, large-calibre pedicle. It is the standard for intra-oral reconstruction, where the tissue's thinness and mobility matter more than its bulk. The donor scar on the forearm is its drawback.
The free gracilis muscle, transferred with its motor nerve to a target nerve in the face, is the standard for in long-standing complete facial paralysis. Harii's 1976 paper introduced the technique25; refinements by Manktelow, Zuker, and others have made it the dominant operation for restoring a smile when the native facial nerve cannot be reused.
A toe-to-hand transfer — Cobbett's 1969 case, evolved into Morrison's wrap-around flap and Wei's trimmed great-toe transfer — replaces a missing thumb with a vascularised, sensate, mobile digit. It remains one of the more visible applications of the principle: a foot is a foot, and a hand without a thumb has lost roughly half its function; the operation gives back what a prosthesis cannot.

Less common but illustrative

Beyond the workhorses, a handful of flaps illustrate where the field has gone.
The , described by Koshima in 200426, takes the same groin territory as Daniel and Taylor's original 1973 flap, but raises only the skin and fat on a single small perforator using supermicrosurgical technique. The donor scar hides in the bikini line. Where a 1973 free flap was a demanding operation requiring centimetre-scale vessels, a 2024 SCIP flap is routinely raised on a 0.8-millimetre artery.
takes a small group of lymph nodes — typically from the groin, the supraclavicular fossa, or the omentum — with their vascular pedicle, and transplants them to a limb affected by lymphedema. The technique was developed by Becker in the 1990s27 and refined by Lin and Cheng28. Outcomes are heterogeneous and patient selection matters; meta-analytic data show clinically meaningful improvement in a substantial proportion of cases29.
— hand and face transplants — is the field's frontier and its most ethically charged application. The first modern hand transplant was performed in Lyon in 199830; the first partial face transplant by the same team in 20053132. Outcomes are recognisable to any transplant clinician: the surgery is the easy part; lifelong immunosuppression and chronic rejection are the difficult ones. VCA is a speciality within the speciality, performed in a small number of centres worldwide.

How well does it work?

Free flap survival in major centres is now reported between 95 and 99 per cent, depending on the centre, the case mix, and the definition of failure3334. The reference baseline is Khouri's 1998 multicentre prospective study, which followed 493 flaps and reported a 4.1 per cent failure rate, identifying irradiated recipient beds and skin-grafted muscle flaps as risk factors33. A more recent single-institution series of 5,000 flaps from the University of Pennsylvania reported success rates above 98 per cent in the contemporary era34.
Failure, when it happens, is usually venous. Around 5–10 per cent of flaps undergo take-back — emergency return to theatre for vascular exploration — and roughly two-thirds of those are salvaged35. Time matters: the diagnosis-to-revision interval is the strongest modifiable predictor of salvage36. This is why postoperative monitoring exists. Clinical observation by a trained nurse remains the foundation; adjuncts include hand-held Doppler, implantable Doppler probes (which improve detection of compromise but require expertise to interpret), and tissue oximetry3738.
The patient-experience figures are different from the surgical ones. For autologous breast reconstruction, large meta-analyses report higher patient-reported satisfaction with breasts and psychosocial well-being for DIEP than for implant reconstruction, with the trade-off of a longer operation, a longer hospital stay, and a donor-site scar3940. For lower-extremity trauma, Marko Godina's 1986 paper established that early reconstruction — within 72 hours — gave dramatically better outcomes than delayed surgery41; modern data extend the safe window to roughly ten days42, but the principle of early definitive cover holds. The British Orthopaedic Association and BAPRAS Standards for the Management of Open Fractures, last revised in 2020, codify the modern orthoplastic pathway43.

Where the field is going

Three directions are visible.
First, supermicrosurgery has become standard practice in lymphatic surgery. connects subdermal lymphatic channels directly to small subdermal venules, providing physiological drainage in patients with lymphedema. Earlier work from Koshima's group established the technique alongside VLNT for selected patients44.
Second, robotic microsurgery has moved from concept to early clinical practice. The MUSA platform was the subject of a 2020 randomised pilot trial in robotic LVA45; the Symani Surgical System, a more recent dedicated microsurgical robot, has been used in early human series for free-flap anastomosis and lymphatic surgery46. Whether robotic assistance ultimately improves on the unassisted human hand at this scale is not yet settled — the current evidence is feasibility-grade — but the trajectory is clear.
Third, imaging has reduced the surgeon's reliance on intra-operative discovery. Pre-operative CT angiography maps perforators before the patient enters theatre. , used intra-operatively, shows perfusion in real time and helps identify flap regions at risk before the wound is closed47.
What none of these advances change is the essential operation: a piece of tissue, lifted on its vascular pedicle, divided, transferred, and reconnected by hand under magnification. Microsurgery rewards practice and patience more than novelty.

A final note

For the patient considering a free flap reconstruction, two things are worth knowing. The operation is long — most free-flap procedures take six to ten hours — and recovery is not trivial: a hospital stay of four to seven days, a return to light activity over four to six weeks, and full recovery measured in months rather than weeks. The trade-off is durable, living tissue that integrates into the body in a way no prosthesis or implant can.
For the referring clinician, the threshold for involving a reconstructive team has lowered as outcomes have improved and pre-operative planning has matured. The decision is rarely whether reconstruction is technically possible; it is whether the patient is fit for the operation and whether the timing is right.
Free tissue transfer is no longer a frontier technique. It is — for many of the defects it addresses — simply the standard of care.