How the US Navy Circumference Method Estimates Body Fat Percentage

The practical question is simple: how close does a tape get to what a DXA scan would find? The answer is a standard error of estimate of ±3.5 percentage points for men and ±3.6 percentage points for women against the original underwater weighing criterion. When the same equation is validated against dual-energy X-ray absorptiometry in general populations, the gap widens to ±3 to ±5 percentage points. That is the margin you carry when a recruiter or a gym coach reads the number from a tape.

The equation is the Hodgdon and Beckett 1984 US Navy tape method, published as Naval Health Research Center Report No. 84-11. It was derived from 602 men and 214 women who were active US Navy personnel aged 18 to 55. The original criterion was hydrostatic weighing, not DXA. The method is a prediction equation that uses logarithms of girth values, not a direct measure of body fat. It gauges subcutaneous and visceral adipose tissue indirectly by the geometry of the torso.

For a man, the equation is: %BF = 86.010 × log₁₀(abdomen girth − neck girth) − 70.041 × log₁₀(height) + 36.76. The abdomen is measured at navel level. The neck is measured just below the larynx. For a woman, the equation is: %BF = 163.205 × log₁₀(abdomen girth + hip girth − neck girth) − 97.684 × log₁₀(height) − 78.387. The abdomen for a woman is taken at the natural waist, the narrowest point. The hip is measured at the largest girth of the buttocks. The height is in inches and the girths in inches, because the equation was built in imperial units.

The standard error of estimate is the key number. Hodgdon and Beckett reported ±3.5% for men and ±3.6% for women against underwater weighing in their original validation sample. That means roughly two-thirds of individuals in that sample had their body fat predicted within 3.5 points of the criterion. The other third were further off. When the same equation is cross-validated against DXA in non-Navy populations, the error range widens. A 1997 literature analysis by Fogelholm et al. found the standard error of estimate across tape-based methods, including the Navy method, runs ±3 to ±5 percentage points when compared to DXA or hydrostatic weighing.

What does that look like in practice? If the Navy equation gives you 18% body fat, the DXA value for the same person on the same day could be anywhere from 13% to 23% and still be within one standard error of the prediction. That is not a small gap. A person at the low end of that range might be considered lean; a person at the high end might be considered overfat. The same tape reading produces both possibilities.

Why the Error Increases Outside the Derivation Population

The equation was derived from US Navy personnel. That is a specific population: young to middle-aged, physically screened for service, with a body composition profile that is not representative of the general public. When you apply the same equation to older adults, to athletes with high muscle mass, to people with obesity, or to ethnic groups whose fat distribution differs from the white European norm, the error widens. The Hodgdon, Beckett equation is a population-specific equation. It works best on people who resemble the 1984 sample. On everyone else, it is a rougher approximation.

The method also has directional bias. Tape-based prediction equations tend to underestimate body fat in very lean individuals and overestimate it in very obese individuals. The geometry of the torso changes at the extremes of body composition, and the log-linear assumption of the equation no longer holds. A lean person may be told they carry more fat than they do; a person with obesity may be told they carry less. Both misclassifications matter in a screening context.

A DXA scan, dual-energy X-ray absorptiometry, distinguishes fat, lean, and bone mass by passing two X-ray energies through the body. Its coefficient of variation for total body fat is 1-3%, making it a reference method for research and clinical settings. The Navy tape method, by contrast, is a surrogate that gauges body composition from the geometry of the torso. The two methods do not measure the same thing. DXA measures the actual attenuation difference between adipose tissue and lean tissue. The Navy equation measures the length of a few girth arcs and plugs them into a formula.

When validation studies compare the two, the error range of ±3-5 percentage points is consistent. That is the standard error of estimate, not the maximum error. Individual outliers can be wider. A 2008 meta-analysis of 25 studies by Romero-Corral et al. found that BMI itself, a cruder tool, has a sensitivity of roughly 50% for identifying excess body fat by DXA. The Navy tape method improves on that because it uses actual body dimensions, but it does not approach the precision of a DXA scan. It is a field-expedient screening tool, not a diagnostic.

When to Trust the Tape and When to Insist on DXA

For a military body composition assessment, the Navy method is the protocol and there is no alternative. For a personal health assessment, the tape is a reasonable starting point if you control the procedure: measure at the correct landmarks, use a non-stretchable tape, take the reading at the end of a normal exhalation, and repeat it three times. Do not use self-measured values. Self-measured waist girth is unreliable, with errors commonly exceeding 3-5 cm, which shifts the body fat approximation by multiple percentage points. If you want a number you can act on for training or clinical decisions, pay for a DXA scan or a skilled skinfold reading by a trained technician using a validated prediction equation.

Variable	Men	Women
Neck measurement site	Just below larynx	Just below larynx
Abdomen measurement site	At navel level	At natural waist (narrowest point)
Hip measurement site	Not used	Largest girth at buttocks
Original criterion method	Hydrostatic weighing	Hydrostatic weighing
Original SEE vs. hydrostatic weighing	±3.5% body fat	±3.6% body fat
SEE range against DXA (validation studies)	±3–5%	±3–5%
Directional bias	Underestimates in lean; overestimates in obese	Underestimates in lean; overestimates in obese
Derivation population	US Navy personnel, n=602	US Navy personnel, n=214

The Hodgdon and Beckett 1984 equation is a prediction equation that uses logarithms of girth differences. The arithmetic is fixed. For a man, the key input is the difference between the abdomen and neck girth. For a woman, it is the sum of the abdomen and hip girths minus the neck girth, divided by height. The log transformation compresses the scale at the high end, which is why the equation tends to compress the approximation at the extremes of body composition.

The equation has not been revised since 1984. The US Navy still uses it for body composition assessment because it is fast, cheap, and non-invasive. It does not require electricity, a certified operator, or a hospital visit. That is its strength. Its weakness is that it was never designed to match DXA. It was designed to match hydrostatic weighing in a population of US Navy personnel. The fact that it is still in use four decades later reflects its practical utility, not its accuracy.

Tape Protocol Error: Why the Person Holding the Tape Matters

The error in the tape method comes from two sources: the equation error and the measurement error. The equation error is the ±3-5 percentage points. The measurement error comes from the tape protocol. If the tape is too tight, the girth is too small and the body fat approximation drops. If the tape is too loose, the girth is too large and the approximation rises. If the landmark is wrong, measuring at the iliac crest instead of the navel, or at the widest part of the hips instead of the largest girth at the buttocks, the error compounds. The inter-observer error for a single girth reading can be 1-3 cm in untrained settings. That shift in a single input changes the body fat approximation by roughly one to two percentage points per centimetre. A measurement error of 2 cm at the waist shifts the body fat approximation by 2-4 percentage points before the equation error is even considered.

To minimise the error: have the same person take all three readings, use a Gulick tape with a spring-loaded mechanism that standardises tension, measure at the correct landmarks as specified in the Hodgdon and Beckett protocol, and take the reading at the end of a normal exhalation. Do not pull the tape tight enough to compress the skin. The tape should sit on the skin but not indent it. If you are measuring yourself, stop. Self-measurement fails because you cannot see the tape at the back, you cannot control the tension consistently, and you tend to hold your breath or flex your abdomen.

The Navy tape method suits the person who needs a repeatable, standardised body fat approximation in a field setting where a DXA scan or a skinfold calliper is not available. It suits the military service member who is being assessed against a service protocol and has no choice of method. It suits the coach who needs a quick, consistent number to track changes in an athlete over time, provided the same person takes every reading at the same time of day under the same conditions.

The Navy equation does not suit the person who wants a precise body fat figure for a clinical decision, a competition weigh-in, or a personal health benchmark. That person needs a DXA scan, a hydrostatic weighing, or a skinfold reading by a trained technician using a validated equation specific to their population and sex. The Navy equation does not suit the person who is outside the derivation population, older than 55, younger than 18, or with a body composition that is not well represented by a military sample. The single thing that most often goes wrong is treating a tape-derived body fat approximation as the truth rather than as a screening number with a known error band.