How to Measure Body Fat at Home

Method 1: Navy tape method

Developed by the U.S. Navy in 1984^[2] as a screening tool for military fitness. Uses only a cloth measuring tape. Inputs: neck, waist, and hip (women only) circumferences, plus height.

Men:    %BF = 86.010 × log10(waist_in − neck_in) − 70.041 × log10(height_in) + 36.76
Women:  %BF = 163.205 × log10(waist_in + hip_in − neck_in) − 97.684 × log10(height_in) − 78.387

The Body Fat Percentage Calculator implements this directly.

Accuracy: ±3–4% body fat against hydrostatic weighing in the original cohort.

Measurement protocol:

• Neck: below the larynx, tape perpendicular to the long axis of the neck.
• Waist: at the navel for men; at the narrowest point between rib cage and hip bone for women. Exhale normally; don't suck in.
• Hip (women): at the widest point of the hips/gluteals, tape horizontal.
• Same time of day each measurement, fasted.

Weakness: tends to overestimate in lean, muscular lifters with narrow waists and large necks (from traps) and heavily-muscled glutes. For lifters at low single-digit body fat, it can read 4–6% high.

Method 2: Skinfold calipers

A €40–€80 pair of skinfold calipers plus the Jackson-Pollock 3-site or 7-site equation^[4]. The 3-site (men: chest, abdomen, thigh; women: triceps, suprailiac, thigh) is easier to self-administer.

Accuracy with a trained tester: ±3–4% body fat. Self-administered on accessible sites: ±5–7%. Longitudinal accuracy (relative change) is substantially better than absolute accuracy — even a crude caliper protocol will show the right direction over weeks.

Technique basics:

• Pinch subcutaneous fat cleanly between thumb and index finger — pull it away from underlying muscle.
• Place calipers 1 cm from the pinch, perpendicular to the fold.
• Wait 2 seconds for caliper reading to stabilise.
• Measure each site three times; use the median.

The most common error is squeezing the calipers harder to get a smaller number. The spring-loaded caliper is supposed to do the work; your job is to place it and read.

Method 3: Consumer BIA scale

Passes AC current through the feet (most home scales) or hand-to-foot (gym machines). Output: body-fat %, often with segmental breakdown.

Accuracy: ±5–8% on consumer foot-to-foot scales^[3]. Hand-to-foot medical-grade devices ±3–4%.

Hydration is the dominant error source. Morning dehydrated = higher BF% reading. Post-dinner or post-workout = lower. A 3% swing on the same day is common and unrelated to actual body fat.

Practical use: same time of day, same hydration status, same day of week. Read the 4-week rolling average, not the single reading.

Method 4: Photos + bodyweight trend

The least technical method and often the most useful for non-competitors. Protocol:

• Standing relaxed, arms at sides, front / side / back photos under consistent lighting and distance.
• Same time of day (morning, fasted, post-bathroom).
• Same pose and angle.
• Every 4 weeks.

Visually comparing four 4-week intervals gives you a clearer picture of actual body-composition change than any single BIA or tape reading. For tracking a recomp, cut, or bulk, photos plus weight trend beats obsessing over a BF% number.

Which method to use

Goal                                   Best method
─────────────────────────────────────────────────────────────────────
Beginner tracking loose weight         Navy tape + photos
Intermediate recomp / slow cut         Navy tape + photos + occasional BIA
Competitive physique prep              Skinfolds biweekly + DEXA baseline
Powerlifter managing weight class      Scale weight, maybe waist circ only
General health                         Photos + fitting clothes

Measurement errors that look like body-composition changes

Several non-composition factors shift at-home body-fat readings:

• Hydration. Post-workout dehydration lowers BIA body-fat estimates by 2–4% while being a water-loss artefact. Salty dinner the night before does the opposite.
• Glycogen. Each gram of glycogen stored binds ~3 g of water. A lifter entering a cut loses glycogen quickly in week 1, producing a 1–2 kg weight drop that's mostly glycogen and water.
• Gut contents. 500–1,500 g of food and fluid in transit across the day.
• Menstrual cycle (women). 1–3 kg water shift across the cycle.
• Travel and sodium changes. Long flights and high-sodium meals produce rapid water retention that shows up as weight gain and misleading body-fat readings.

Measuring under matched conditions (same time, same hydration status, same cycle day) removes these confounds. Measuring at random times with random hydration makes the signal very hard to see.

Combined methods for better signal

You can reduce measurement noise by combining methods:

• Waist circumference + bodyweight. If weight is stable or down while waist is down, body-fat is clearly falling. If weight is stable and waist is stable, you're at maintenance.
• Photos + bodyweight. Visual change is usually visible before tape or skinfold changes.
• BIA trend + Navy tape. If both drift in the same direction over 4 weeks, the signal is robust. If they disagree, one is noise.

What to do with the number

A body-fat number is most useful as a check against extreme values:

• Below ~8% (men) / ~16% (women) for more than a few weeks is associated with hormonal disruption, sleep issues, performance decrement.
• Above ~25% (men) / ~32% (women) is associated with the usual metabolic risk factors.
• In between, the specific number matters less than trend and fitness outcomes.

For a lifter between 10% and 20%, fixating on “am I 12% or 14%?” is not productive. The method error exceeds the resolution you're trying to detect.

Tracking frequency

• Tape method: every 2 weeks.
• Skinfolds: every 2–4 weeks with same tester.
• BIA: weekly, same conditions.
• Photos: every 4 weeks.
• DEXA: every 3–6 months.

More frequent measurement increases noise without adding signal. Daily BIA readings are the worst of both worlds.

Navy tape (waist 83, neck 38.5, height 173):  14.8% BF
Jackson-Pollock 3-site skinfolds:             13.2% BF
Consumer BIA scale (4-day mean):              15.9% BF
Derived range (±3% method error per source):  12–17% BF