Which BSA Formula? Mosteller vs Du Bois vs Haycock

Q: Which BSA formula is the most accurate?

None has been shown to be clearly superior across all body sizes. Each was derived from a small group of directly measured people, so accuracy depends on how closely a patient resembles that group. In everyday practice, using the protocol’s formula consistently matters more than the choice itself.

Q: Why is Mosteller used so widely?

It needs only a multiplication, a division and a square root, so it can be done on any calculator, and it tracks the more complex equations closely in adults of average build. It is the default on the DoseStep BSA calculator for the same reason. In obesity the formulas drift apart, so the choice matters more there.

Q: How much do the formulas differ?

For a 70 kg, 170 cm adult the four results fall between 1.81 and 1.83 m², about 1%. For a 150 kg, 170 cm adult they range from 2.50 to 2.75 m², a spread of about 10%. At 75 mg/m², that moves the dose from 187.5 to 206.25 mg.

Q: Should BSA be capped at 2.0 m² in obesity?

Capping was once common but is now debated. The ASCO guideline on chemotherapy dosing in obese adults (2012, updated 2021) recommends full weight-based doses using actual body weight, particularly when the aim is cure. Some protocols still cap BSA for specific drugs, and practice varies between centres.

Q: Is BSA used for children?

Yes. Haycock was developed and validated from premature infants to adults, and Mosteller is also widely used in paediatrics. In very small children the equations differ by a larger percentage, even though the absolute gap in m² is small: for a 3.5 kg newborn, the four results run from 0.21 to 0.23 m².

Q: Does BSA replace mg/kg dosing?

No. They are different prescribing bases. A drug is dosed per kg, per m², as a fixed dose or by another method such as target exposure (carboplatin by AUC, for example). The basis comes from the product information or the protocol, not from the calculator.

At a glance

Formula	Year	Equation (kg, cm → m²)	Derived from	Typical use
Du Bois	1916	0.007184 × kg^0.425 × cm^0.725	9 people, measured by moulds	Historic reference; physiology, some oncology
Gehan & George	1970	0.0235 × kg^0.51456 × cm^0.42246	About 400 earlier direct measurements	Oncology (mainly US)
Haycock	1978	0.024265 × kg^0.5378 × cm^0.3964	81 people, premature infants to adults	Paediatrics
Mosteller	1987	√(kg × cm ÷ 3600)	Simplification fitted to existing data	Bedside default in many settings

In short: for adults of ordinary build the choice changes the result by up to about 2%. In obesity and in newborns the spread grows to around 10%, and Du Bois tends to give the lowest value in heavy adults.

Where each BSA formula comes from

Du Bois (1916). Delafield and Eugene Du Bois coated nine subjects in thin moulds, measured the surface and fitted an equation to height and weight. Only one subject was a child, and nine people cannot represent the range of body shapes seen in practice. Despite this tiny sample, it remained the standard reference for most of the twentieth century and still appears in many physiology texts.

Gehan & George (1970). Edmund Gehan and Stephen George re-analysed about 400 direct surface measurements collected by earlier researchers and produced a statistically fitted equation. It became common in oncology, particularly in North American cooperative trials.

Haycock (1978). George Haycock and colleagues measured body surface geometrically, treating the body as a set of cylinders and other shapes, in 81 people from premature infants to adults. Because the sample covered the whole paediatric range, it is often preferred in children.

Mosteller (1987). Raymond Mosteller proposed, in a short letter, a simplified formula that reproduces the older equations closely while needing only a square root. Its simplicity made it the default in many hospitals, in electronic prescribing systems and on the DoseStep BSA calculator.

How much do they differ?

The table below uses exactly the four equations in the site’s calculator, with results rounded to two decimals, as the calculator displays them.

Weight, height	Mosteller	Du Bois	Haycock	Gehan & George	Spread
3.5 kg, 50 cm (newborn)	0.22	0.21	0.22	0.23	0.025 m² (12%)
20 kg, 115 cm (child)	0.80	0.80	0.80	0.81	0.018 m² (2%)
50 kg, 155 cm	1.47	1.47	1.47	1.48	0.014 m² (1%)
70 kg, 170 cm	1.82	1.81	1.83	1.83	0.022 m² (1%)
90 kg, 190 cm	2.18	2.18	2.18	2.18	0.005 m² (< 1%)
110 kg, 165 cm	2.25	2.15	2.30	2.28	0.155 m² (7%)
150 kg, 170 cm	2.66	2.50	2.75	2.71	0.249 m² (10%)

Three patterns stand out:

Average adults: all four agree within about 0.04 m². The choice rarely changes a dose in a meaningful way.
Heavy, short adults: the equations diverge. Du Bois gives the lowest value because its weight exponent (0.425) is the smallest; Haycock gives the highest.
Newborns: the absolute gap is tiny, but in relative terms it is the largest in the table, because the equations were fitted mostly on bigger bodies.

BSA-based dosing (mg/m²)

Some medicines are prescribed per square metre, most often cytotoxic chemotherapy, because BSA was thought to track metabolic rate and drug clearance better than weight. The calculation itself is simple:

Dose (mg) = dose per m² × BSA (m²)

Example: 75 mg/m² for a 70 kg, 170 cm adult. Mosteller BSA = √(70 × 170 ÷ 3600) = 1.82 m². Dose = 75 × 1.82 = 136.5 mg. Using Du Bois (1.81 m²) instead gives 135.75 mg, a difference of under 1%.

Example: the same 75 mg/m² for a 150 kg, 170 cm adult.

Approach	BSA (m²)	Dose
Mosteller	2.66	199.5 mg
Du Bois	2.50	187.5 mg
Haycock	2.75	206.25 mg
BSA capped at 2.0 m²	2.00	150 mg

Here the formula alone moves the dose by almost 19 mg, and a 2.0 m² cap reduces the Mosteller dose by about a quarter. That is why the obese patient is where the choice of equation, and the question of capping, actually matters.

BSA does not suit every drug. Some medicines have an absolute maximum per dose regardless of size, and others are dosed by target exposure (for example carboplatin by AUC with a kidney-function estimate) rather than by m². For weight-based orders, see the mg/kg dose calculator.

Capping BSA in obesity: the debate

For decades many centres capped BSA at 2.0 m², or dosed on ideal or adjusted weight, fearing toxicity in larger patients. The 2012 American Society of Clinical Oncology (ASCO) guideline reviewed the evidence and reached a different conclusion:

Full weight-based chemotherapy doses, using actual body weight, were recommended for obese adults, especially when treatment aims at cure.
The panel found no consistent evidence that obese patients had more toxicity with full doses, and noted that routine reduction might compromise outcomes.
Any of the standard BSA formulas was considered acceptable.
If toxicity occurs, dose reductions should follow the same rules as for non-obese patients, rather than a systematic up-front cut.

ASCO updated the guideline in 2021, extending the same principle to other systemic therapies. Even so, capping persists in some protocols, for specific drugs, or in particular clinical situations, and practice varies between countries and institutions. We recommend following the protocol and prescription that apply to your patient.

If an adjusted or ideal weight is used by a protocol, the ideal body weight and adjusted body weight calculators show the intermediate steps. The guide to actual, ideal and adjusted body weight explains when each one is used.

Common pitfalls

Units. All four equations expect kilograms and centimetres. Entering pounds or inches, or height in metres, gives a meaningless result.
Mixing formulas. Calculating the cycle 1 dose with one formula and the next cycle with another can shift the dose without any change in the patient.
Stale weight. BSA depends on current weight; a weight from weeks ago may no longer reflect the patient, especially during treatment that causes weight loss.
Rounding too early. Rounding BSA to one decimal (1.8 instead of 1.82) changes a 75 mg/m² dose by 1.5 mg; dose banding, where used, is a separate, deliberate step defined by the protocol.
Assuming a cap. Whether to cap at 2.0 m² is a prescribing decision, not a calculator default. The DoseStep calculator never caps.
Over-reading small differences. In average adults, a 0.01–0.02 m² gap between formulas is well below the uncertainty of the equations themselves.

Measuring height and weight

Head straight, looking ahead, with the headboard resting on the top of the head.
Barefoot, heels together and against the wall or the measuring rod.
Scales on a hard, flat floor, in light clothing; weighing at the same time of day makes changes easier to compare.

Frequently asked questions

Which BSA formula is the most accurate?

None has been shown to be clearly superior across all body sizes. Each was derived from a small group of directly measured people, so accuracy depends on how closely a patient resembles that group. In everyday practice, using the protocol’s formula consistently matters more than the choice itself.

Why is Mosteller used so widely?

It needs only a multiplication, a division and a square root, so it can be done on any calculator, and it tracks the more complex equations closely in adults of average build. It is the default on the DoseStep BSA calculator for the same reason. In obesity the formulas drift apart, so the choice matters more there.

How much do the formulas differ?

For a 70 kg, 170 cm adult the four results fall between 1.81 and 1.83 m², about 1%. For a 150 kg, 170 cm adult they range from 2.50 to 2.75 m², a spread of about 10%. At 75 mg/m², that moves the dose from 187.5 to 206.25 mg.

Should BSA be capped at 2.0 m² in obesity?