How to Measure Body Fat Percentage (And Which Numbers to Trust)

A practical guide to measuring body fat with tape, calipers, bioimpedance and DEXA, how far each method misses, and how to keep your numbers comparable week to week.

The scale tells you how heavy you are, not what you lost — fat, water or muscle. Body fat percentage fills that gap, but only if you know how far each method misses and you measure the same way every time. Here is a workable process, no lab and no expensive gear required.

What you are actually measuring

Body fat percentage is the share of fat tissue in your total mass. At 90 kg and 20% fat, that is 18 kg of fat and 72 kg of lean mass — muscle, bone, water, organs. The goal of a fat loss phase is to take those 18 kg down to, say, 12 kg while the 72 kg stays untouched.

No device measures fat directly. They all measure something else — skinfold thickness, electrical resistance, body density — and then run it through an equation. That is why the absolute number is almost always off, and why the difference between two measurements taken the same way is the part that matters.

The methods and their error

Tape measure

Cheapest and surprisingly useful. Measure waist at the navel, hips at the widest point, neck below the Adam's apple, and optionally upper arm and thigh. The tape should sit flush against skin without denting it. Take three readings and use the middle one; the spread should stay under 0.5 cm.

Waist circumference carries health information on its own: over 94 cm in men means raised risk and over 102 cm high risk; in women over 80 cm raised and over 88 cm high. An even simpler rule — waist divided by height should stay under 0.5. At 180 cm tall, that means a waist under 90 cm.

Skinfold calipers

A clamp that measures the thickness of a pinched fold. The standard three sites are chest, abdomen and thigh for men; triceps, thigh and the fold above the hip for women. Take each site two or three times, wait 2 seconds after closing the caliper, use the middle reading, and feed the numbers into a Jackson-Pollock formula.

Expect about 3 to 4 percentage points of error when the same person measures you every time, and considerably more when the tester changes. Always use the right side of the body and always measure before training, not after.

Bioimpedance

Smart scales and handheld units send a weak current through the body and estimate fat from resistance. Error runs 4 to 8 percentage points, and the reading swings 2 to 3 points from hydration, a salty dinner, yesterday's session or a full bladder alone. That number is only meaningful as a series taken under identical conditions.

DEXA and BodPod

Lab methods, with roughly 1 to 2 percentage points of error. DEXA also shows fat distribution and bone density. Worth doing once or twice a year as a reference point, not as a monthly check.

The protocol that makes numbers comparable

Measure in the morning, within 30 minutes of waking, after the bathroom, before food, coffee and training.
Same day of the week, same device, same sites, same person holding the calipers.
No alcohol and no unusually salty food for 24 hours beforehand.
Log three things together: weight, circumferences and the fat estimate. None of them tells the full story alone.
Draw no conclusions before 4 measurements, meaning 4 weeks. You are reading a trend, not a data point.

For women: the cycle shifts water retention by up to 1.5 kg. Measure in the same phase every time, for example days 7 to 10 after your period starts.

What counts as a realistic number

Men: essential fat 3–5%, athletic 6–13%, fit 14–17%, average 18–24%, obese above 25%.
Women: essential fat 10–13%, athletic 14–20%, fit 21–24%, average 25–31%, obese above 32%.

Below 6% for men and below 14% for women is not a state you hold without a bill coming due — sleep, libido, strength and immune function all drop. Those are numbers for a few weeks around a contest, not for the whole year.

What to do with the number once you have it

A realistic rate is 0.5 to 1% of body mass per week. At 90 kg that is 0.45 to 0.9 kg. In practice: a 300–500 kcal daily deficit, 1.6–2.2 g of protein per kilogram of body mass (145–200 g at 90 kg), 2 to 4 strength sessions a week with 10 to 20 working sets per muscle group, and 7,000–9,000 steps a day as a baseline.

The check: if weight drops but your waist has not moved in four weeks, you are most likely losing water and muscle rather than fat — the deficit is too aggressive, or protein and resistance training are too low.

When to see a doctor

Unintentional loss of more than 5% of body mass in 6 months with no change in diet or training.
Waist over 102 cm in men or 88 cm in women — worth checking fasting glucose, HbA1c, a lipid panel and liver enzymes.
Missed periods for 3 months or longer, especially alongside low body fat and a heavy training load.
Persistent fatigue, feeling cold, hair shedding and falling strength despite decent sleep and food — get thyroid and iron checked.
Compulsive weighing, fear of meals, or cycles of restriction and bingeing. That belongs with a doctor or therapist, not a new diet.

A method you will repeat 20 times the same way beats a precise method you use once.

The short version

No method gives a true number — you are tracking a trend, not an absolute value.
Tape and waist circumference are the cheapest and the most clinically useful; waist under half your height is a solid target.
Calipers miss by 3–4%, bioimpedance by 4–8%, DEXA by 1–2%; pick based on how often you plan to measure.
Morning, fasted, same weekday, same device, minimum 4 weeks before you judge anything.
Aim for 0.5–1% of body mass per week with 1.6–2.2 g protein per kilogram and 2–4 strength sessions weekly.