Body Surface Area Explained: The Five Formulas and Why Clinicians Use mg/m²

What is body surface area?

Body surface area, or BSA, is the total external area of skin covering a human body. It is reported in square metres, and for adults it almost always sits between 1.5 and 2.2 m². The figure matters in clinical medicine for one principal reason: many physiological processes — basal metabolic rate, glomerular filtration, cardiac output, drug distribution — scale more reliably with surface area than with body mass. If you have ever been told a chemotherapy dose “in milligrams per metre squared,” that metre squared is BSA.

Direct measurement is impractical — the historical method involved wrapping a subject in paper or fabric, cutting the moulded shape free, and measuring the area — so clinical BSA is computed from height and weight using one of five published formulas. The body surface area calculator on this page implements all five (Mosteller, DuBois & DuBois, Haycock, Gehan & George, and Boyd) so you can see how much they agree on your numbers. For most adults they fall within about one and a half per cent of each other. At the extremes of weight, particularly in young children and people carrying large amounts of body fat, the gap widens.

BSA is not body composition. It tells you nothing about how much of you is fat versus lean tissue, nothing about hydration, nothing about fitness. It is a geometric estimate that turns out to be useful clinically because area scales with physiological capacity in ways that mass does not.

How BSA is calculated

Five formulas dominate the literature. They were fitted on different populations between 1916 and 1987, and they all take the same two inputs — height and weight — but pair them up under different exponents.

Mosteller (1987)

BSA = √(H × W / 3600)

H is height in centimetres, W is weight in kilograms. Mosteller published this simplified form in a letter to the New England Journal of Medicine after noticing that the older DuBois formula, while accurate, was awkward at the bedside. The square root is biologically motivated: surface area scales as the square of a linear dimension while mass scales as the cube, so the product of a linear (height) and a mass-like (weight) quantity under a square root captures the geometry of a body of roughly constant density. The 3600 in the divisor is the regression constant that aligns the fit to DuBois. Mosteller is the modern clinical default.

DuBois & DuBois (1916)

BSA = 0.007184 × W0.425 × H0.725

The original. Delbert and Eugene DuBois coated nine subjects in moulding paper at Cornell, measured the surface area directly, and regressed against height and weight to derive the power-law fit. With a sample of nine the formula is amazingly durable — it agrees with modern fits within about three per cent across the adult range. Most older clinical nomograms, including the foundational chemotherapy dosing tables, are built on DuBois.

Haycock (1978)

BSA = 0.024265 × W0.5378 × H0.3964

Haycock and colleagues re-fitted the formula on a sample that included infants and small children, which DuBois’s original sample did not. Below about 50 kg Haycock gives meaningfully different (and more accurate) BSAs than the original DuBois fit, which is why paediatric oncology and nephrology services use it preferentially. Above 50 kg Haycock and Mosteller are almost indistinguishable.

Gehan & George (1970)

BSA = 0.0235 × W0.51456 × H0.42246

A 401-subject re-fit that included a broader range of body sizes than DuBois had access to. Gehan and George argued that a single power law could not capture the relationship across all ages and so re-derived the coefficients with a much larger dataset. Their formula is used in some oncology protocols and in burn-care literature, but it has never displaced Mosteller as the bedside default.

Boyd (1935)

Boyd’s formula is the odd one out: instead of fixed exponents it uses a log-corrected weight exponent that varies with the weight itself. In its commonest written form,

BSA = 0.0003207 × H0.3 × Wg(0.7285 − 0.0188 × log10 Wg)

where W_g is weight in grams. The variable exponent gives Boyd a slight advantage at very high body weights, where a fixed power law begins to over- or under-predict. Most modern clinicians never see Boyd, but it is included in the BSA calculator breakdown for completeness.

Worked example

Take an adult who is 178 cm tall (5 ft 10 in) and weighs 84 kg (185 lb). Running all five formulas:

• Mosteller: √(178 × 84 / 3600) = √4.1533 = 2.038 m²
• DuBois: 0.007184 × 84^0.425 × 178^0.725 ≈ 2.022 m²
• Haycock: 0.024265 × 84^0.5378 × 178^0.3964 ≈ 2.051 m²
• Gehan & George: 0.0235 × 84^0.51456 × 178^0.42246 ≈ 2.045 m²
• Boyd: with weight in grams, the variable exponent works out at about 0.6816, giving 2.034 m²

All five values fall inside a 1.5 per cent band. For a typical adult that is the level of agreement you can expect — and is why arguments about which formula is “best” are usually beside the point. Most clinicians round to two decimal places and use the Mosteller value (2.04 m² here) for dosing.

For a paediatric example, take a child of 100 cm and 30 kg. Mosteller gives √(100 × 30 / 3600) = 0.913 m². Haycock — the formula fitted on children — gives 0.938 m². That three per cent gap is the paediatric correction Haycock was designed to capture, and it is exactly the kind of difference that nudges a clinician toward Haycock when dosing a small child. Run both through the BSA calculator and the spread is visible in the breakdown.

Why clinicians use BSA instead of weight

Drugs that distribute through the bloodstream and are cleared by the kidneys or liver follow pharmacokinetics that track organ capacity. Organ mass and blood volume both scale with body surface area more reliably than with body weight, particularly across a heterogeneous patient population. Two adults of the same weight but very different heights have meaningfully different blood volumes and very different clearance rates for many drugs. They have closer BSAs, and therefore closer mg/m² doses produce closer drug exposures.

Cytotoxic chemotherapy is the canonical case. The therapeutic window between “effective” and “toxic” is narrow for almost all chemotherapy agents, and the consequences of under-dosing (treatment failure) and over-dosing (severe haematological toxicity, sometimes lethal) are both serious. Dosing in mg/m² rather than mg/kg reduces the inter-patient variability in drug exposure by roughly a third for many regimens, which is the difference between a tolerable regimen and an intolerable one. The same argument applies, though less acutely, to many anaesthetic agents, biologics dosed on body size, and renal-replacement prescriptions where dialysate clearance is set per m².

BSA is not a panacea. A growing body of pharmacokinetic research suggests that for some monoclonal antibodies and small-molecule targeted therapies, flat dosing or weight-based dosing performs at least as well as BSA-based dosing. The field has not abandoned mg/m² — it is too entrenched in the literature, the labelling, and the clinician’s mental model — but for newer agents the dosing rationale is reviewed on a per-drug basis.

Which formula to use when

Adult clinical use: Mosteller

Mosteller is the modern default. It is the formula most chemotherapy nomograms are calibrated against, it is trivial to compute by hand or at the bedside, and across the normal adult range it agrees with DuBois to within about one per cent. Unless a specific protocol mandates otherwise, Mosteller is the right answer for an adult.

Paediatric use: Haycock

Paediatric oncology and nephrology services typically use Haycock, because it was fitted on a population that included children and gives more accurate BSAs below about 50 kg. Some paediatric services use Mosteller across the full weight range for simplicity; the resulting doses are slightly lower in small children, which is usually acceptable but worth flagging if a protocol assumes Haycock.

Historical comparisons: DuBois

Older textbooks, nomograms, and case reports often quote DuBois values. If you are reconciling a calculation against a published reference from before about 1990, use DuBois so the comparison is apples-to-apples. For new calculations from scratch there is no clinical reason to prefer DuBois over Mosteller.

Specialised protocols: Gehan & George or Boyd

A handful of oncology protocols and burn-extent estimates specify Gehan & George. Boyd appears occasionally in nutrition and exercise physiology literature, particularly where investigators want a formula that performs slightly better at extreme body weights. If a protocol explicitly names a formula, use the one named — do not substitute Mosteller without a clinical reason.

The chemotherapy cap and other limitations

Most chemotherapy protocols cap the BSA used for dosing at 2.0 m², regardless of the patient’s actual computed BSA. The reasoning is empirical: the linear dose response that justifies mg/m² dosing was derived on patients within the normal BSA range, and extrapolating a linear relationship to very large patients is a known way to over-dose them. Some services use a 2.2 m² cap, and practice varies by institution and by drug. The cap is a convention, not a derivation — if you see a quoted chemotherapy dose that looks lower than the raw BSA would suggest, the cap is the likely explanation.

BSA computed from height and weight is an estimate. Direct measurements agree with computed BSAs to within about five per cent for adults of normal build, and the disagreement widens at the extremes. For very obese adults, very lean athletes, or amputees, none of the five formulas can be expected to be precisely right. In those cases the BSA is used as the best available estimate and the dosing decision is informed by clinical judgement, not just by the number.

BSA also does not adjust for body composition. Two adults with the same height and weight have the same computed BSA, even if one is a bodybuilder and the other is sedentary. For anything where composition matters — lean drug clearance, exercise prescription, nutritional planning — pair the BSA reading with a body-composition estimate such as the lean body mass calculator or the BMI calculator.

Common mistakes

Treating Mosteller and DuBois as interchangeable in children

In adults the two formulas agree to within one per cent and can be used interchangeably. In small children Mosteller systematically under-predicts compared to Haycock or DuBois, and the gap can reach five per cent or more. Paediatric dosing decisions made on a Mosteller BSA can therefore be meaningfully lower than the same decision made on Haycock. Stick to one formula throughout a course of treatment.

Mixing unit systems mid-calculation

Every BSA formula expects specific units — height in centimetres, weight in kilograms, with the single exception of Boyd which uses grams for weight. Plug pounds into a kilogram formula and the result is off by a factor of about 2.2. The BSA calculator converts internally with exact NIST factors, but if you are doing the algebra by hand, convert first using the weight converter and the height fields in cm. Pre-convert before you start the formula.

Reading the third decimal place

A reported BSA of 1.957 m² is no more precise than 1.96 m². The underlying formulas are accurate to about five per cent at best, and the height and weight inputs are usually rounded to the nearest centimetre and the nearest half-kilogram. Two decimal places is plenty; three is false precision.

Quoting a chemotherapy dose without the cap

If you are sense-checking a chemotherapy regimen against a published mg/m² dose, remember that most protocols cap BSA at 2.0 (or 2.2) m². A patient with a raw BSA of 2.4 will usually be dosed as if their BSA were 2.0, not 2.4. The gap is not an error in the calculator; it is an institution-level convention applied on top of the formula.

When to seek professional advice

The BSA on its own is a number, not a clinical decision. Situations where you should not rely on a self-computed value include:

• Planning or interpreting a chemotherapy or biologic dose. Oncology pharmacies recompute BSA at each cycle and apply institutional capping rules; the dosing decision belongs with the prescriber and pharmacist, not the patient.
• Paediatric drug dosing of any kind. Children’s BSAs change quickly with growth and a stale value can be materially wrong. Paediatric services recompute at every visit.
• Burn-extent assessments in trauma. The Wallace “rule of nines” estimates burn area as a percentage of total body surface area, but the computed BSA is one input to a clinical assessment, not a standalone diagnosis.
• Renal-replacement prescriptions, where dialysate clearance is set per m² and is best computed against measured values rather than estimates in patients with extreme body composition.

Frequently asked questions

Which BSA formula should I use? For adult clinical use, Mosteller. It is the bedside default, agrees with DuBois to within one per cent across the normal adult range, and is the formula most chemotherapy nomograms are built around. For paediatric patients, Haycock — it was fitted on a population that included children and is more accurate below 50 kg. DuBois remains the historical reference for comparisons against pre-1990 literature.

Why is BSA used for drug dosing instead of body weight? Many physiological processes — basal metabolic rate, glomerular filtration rate, cardiac output, blood volume — scale more closely with surface area than with mass. Cytotoxic chemotherapy in particular has narrow therapeutic windows, and dosing in mg/m² reduces inter-patient variability in drug exposure compared with mg/kg. Some newer agents are moving back to flat or weight-based dosing where the pharmacokinetics support it, but mg/m² remains the default for most cytotoxics.

What is a normal body surface area? The average adult BSA is about 1.7 m² for women and 1.9 m² for men. The full normal adult range is roughly 1.5 to 2.2 m². A newborn is around 0.25 m², a one-year-old around 0.45 m², a five-year-old around 0.75 m². Very large adults can exceed 2.5 m², though chemotherapy dosing is usually capped at 2.0 m² regardless.

How accurate are computed BSAs? Computed BSAs agree with direct measurement to within about five per cent for adults of normal build. They are less accurate at the extremes — very obese, very thin, or very young patients can have measured BSAs that differ from computed by ten per cent or more. For most clinical purposes that error is dwarfed by inter-patient pharmacokinetic variability, so any of the five published formulas is fit for purpose.

Why does Mosteller use exactly 3600 in the divisor? It is a regression constant, not a biological quantity. Mosteller fitted a square-root model to the same data DuBois used, and 3600 is the constant that makes the square-root fit closest to DuBois across the adult range. The square-root form itself is biologically motivated — area scales as the square of a linear dimension while mass scales as the cube — but the number 3600 is the artefact of a regression, not a derivation.

Is BSA the same as BMI? No. BMI is a ratio of weight to height squared used as a rough proxy for adiposity; BSA is an estimate of the actual external surface area of the body in square metres. BMI tells you whether someone is underweight, normal, overweight or obese for their height; BSA tells you a physiologically meaningful absolute quantity that scales with metabolic capacity. A 178 cm 84 kg adult has BMI ≈ 26.5 and BSA ≈ 2.04 m². Different numbers, different purposes — you can run the BMI half on the BMI calculator.

Does the BSA calculator support imperial units? Yes. Enter height in centimetres or inches and weight in kilograms or pounds; the BSA calculator converts both to SI internally using exact NIST factors (1 inch = 2.54 cm; 1 lb = 0.45359237 kg) before applying the formula. The result is identical to entering metric values directly, to within rounding of the displayed inputs.