Why a tape measure can estimate fat

Fat has to go somewhere. In men it accumulates first around the abdomen; in women, around the hips and thighs as well. Because the trunk and neck change shape as fat is laid down while the limbs change comparatively little, a few circumferences carry real information about composition — enough that the United States Navy adopted a circumference protocol as a cheap alternative to calipers and immersion weighing.

The method's appeal is equipment: a flexible tape, no training beyond consistent landmarking, and no radiation, cost or appointment. Its weakness is that it infers composition from geometry, so it is an estimate with a genuine margin, not a reading.

The equations this page implements

Both formulas are logarithmic and the coefficients are calibrated for inches. The tool converts every circumference to inches before applying them, so metric and imperial input give the same answer — 172 cm is evaluated as 67.717 in, not as 172. Feeding the equation centimetres instead is a common error and inflates the result badly, as the worked example below shows.

male:   %fat = 86.010 × log10(waist − neck) − 70.041 × log10(height) + 36.76

female: %fat = 163.205 × log10(waist + hip − neck) − 97.684 × log10(height) − 78.387

Men supply height, waist and neck; women supply height, waist, neck and hip, and the hip field appears only after the sex selector is set. The tool clamps the output to the range 2–60 per cent so that a mistyped circumference cannot return an impossible figure, and it reports lean mass as 100 − %fat, which is the share of body mass that is not fat.

The result is then placed in one of five bands. For men: essential 2–5, athletic 6–13, fit 14–17, average 18–24, high 25 and over. For women the same bands sit higher: essential 10–13, athletic 14–20, fit 21–24, average 25–31, high 32 and over. The difference is not a judgement about fitness; it reflects the fat required for hormonal and reproductive function.

Walking through the male defaults

Height 172 cm, waist 82 cm, neck 38 cm. Converting to inches first: 172 ÷ 2.54 = 67.717, 82 ÷ 2.54 = 32.283, 38 ÷ 2.54 = 14.961.

1. waist − neck = 32.283 − 14.961 = 17.323
2. log10(17.323) = 1.23862, so the first term is 86.010 × 1.23862 = 106.532
3. log10(67.717) = 1.83069, so the second term is 70.041 × 1.83069 = 128.223
4. 106.532 − 128.223 + 36.76 = 15.1 per cent, band Fit
5. Lean mass: 100 − 15.1 = 84.9 per cent

The default female case adds the hip term. Height 165 cm, waist 72 cm, neck 32 cm, hip 96 cm — in inches 64.961, 28.347, 12.598 and 37.795:

1. waist + hip − neck = 28.347 + 37.795 − 12.598 = 53.543
2. 163.205 × log10(53.543) = 163.205 × 1.72871 = 282.131
3. 97.684 × log10(64.961) = 97.684 × 1.81265 = 177.067
4. 282.131 − 177.067 − 78.387 = 26.7 per cent, band Average

Where this method goes wrong

• Many other calculators get the units wrong. These coefficients are calibrated for inches. Applying them directly to centimetres — which plenty of online calculators do — inflates the male example above from 15.1 to about 21.5 per cent, a six-point error that looks entirely plausible. This tool converts to inches first, so its figure follows the published equation. If another calculator disagrees with it by roughly that much, check which units that one is using rather than assuming either is broken.
• The tape is the dominant error. Differentiating the male formula shows the result moves about 0.85 percentage points per centimetre of waist or neck error, and only about 0.18 points per centimetre of height. A tape pulled tight instead of snug, or a neck measured above rather than below the larynx, is worth several points on its own.
• Measurement sites are specified and easy to get wrong. Men measure the waist at the navel; women at the narrowest point. The hip is taken at the widest part of the buttocks, and the neck just below the larynx, sloping slightly downward at the front.
• Very lean and very heavy bodies sit outside the fitted range. The clamp at 2 and 60 hides that rather than fixing it.
• Logarithms make the formula unforgiving near the inputs. If waist ever fails to exceed the neck by a sensible margin, the first term collapses or goes negative and the output is meaningless.
• None of this is a medical measurement. DEXA and hydrostatic weighing are the reference methods; this is a field estimate.

Common questions

Is the lean-mass figure in kilograms?

No, and this trips people up. It is the complementary percentage, not a mass. Multiply it by your body weight to get lean kilograms.

Why did my two readings move two points in a week?

Because a centimetre of measurement difference is worth nearly a point. Take three readings a week and average them; a real change takes weeks and is visible in the average, not in a single morning.

Should I trust this over my smart scale?

Neither is a reference method. Bioimpedance scales are strongly affected by hydration, so a hard training session or a salty meal moves them several points. Circumference methods move with the tape. Pick one and track it.

Do the bands differ for athletes?

They are the same numbers read differently. A trained male sprinter at 12 per cent sits in the athletic band; the same figure in an untrained person means something else, because the equation cannot see where the lean mass came from.

How does this relate to BMI?

BMI fixes the weight-for-height question and ignores composition; this page attacks composition directly and ignores weight. A muscular person can score as overweight on the BMI page and healthy here, which is exactly why the two pages are best read together.