A tape measure supplies an estimate, not a direct fat measurement
You can estimate body fat percentage from height and selected body circumferences using historical equations commonly called the Navy tape method. A calculator combines those measurements into a predicted fraction of body weight. It does not observe fat tissue or establish your true percentage. The distinction matters when searching for how to measure body fat at home: the tape measures lengths, while the equation supplies an estimate based on a population model.
NumUtility's Body Fat Calculator uses the historical male and female circumference equations with consistent unit conversion. Its three outputs are estimated body fat percentage, estimated fat mass, and estimated fat-free mass. It does not calculate a healthy category, ideal percentage, or personal target. This guide explains the method for educational purposes and is not medical advice. It is not intended for children, pregnancy, breastfeeding, competitive athletes, or circumstances that require individual clinical interpretation.
Use the measurement sites that belong to the selected equation
The historical male term uses standing height, neck circumference, and abdomen circumference at the navel. The female term uses standing height, neck circumference, the narrowest natural waist, and hip circumference at the widest part of the buttocks. These waist and abdomen sites are not interchangeable. A convenient belt position or the waist dimension on a clothing label is not a substitute for the specified circumference. Record the site as well as the number.
Use a nonstretch measuring tape, keep it against the body without compressing the skin, and measure the abdomen or waist with a relaxed posture after a normal exhalation. Place the neck measurement below the larynx rather than across it. Check that the tape is level around waist and hips. These are practical measurement checks, not a complete military assessment protocol. The binary equation terms are limitations of the historical model and do not describe every person's physiology.
Keep every equation length in inches
For the male equation, P = 86.010 × log10(A − N) − 70.041 × log10(H) + 36.76. For the female equation, P = 163.205 × log10(W + B − N) − 97.684 × log10(H) − 78.387. Here P is the predicted percentage, H is height, N is neck, A is abdomen at the navel, W is natural waist, and B is hip circumference. All lengths inside these particular equations are inches.
Metric input must be divided by 2.54 before applying these constants. Do not insert centimeters into an inches-based equation or mix the constants with a separate body-density formula found elsewhere. The function is log10, not the natural logarithm ln. The circumference combination must also be positive before taking a logarithm. An error here is an arithmetic-domain or input problem, not evidence that a negative body-fat value should be rounded up to zero.
Worked male example: percentage and mass are separate steps
Consider a hypothetical adult aged 40 with height 70 in, neck 15 in, abdomen 34 in, and weight 78 kg. The circumference difference is 34 − 15 = 19 in. The equation is 86.010 × log10(19) − 70.041 × log10(70) + 36.76, approximately 17.5130854%. The displayed estimate is 17.5%. Equivalent lengths are height 177.8 cm, neck 38.1 cm, and abdomen 86.36 cm, which must produce the same percentage.
The estimated fat fraction is the unrounded percentage divided by 100. Multiplying 78 kg by 0.175130854 gives approximately 13.6602 kg fat mass, displayed as 13.7 kg. Subtracting that unrounded amount from 78 gives approximately 64.3398 kg fat-free mass, displayed as 64.3 kg. Do not multiply weight by the rounded 17.5% when checking the calculator's internal arithmetic. These illustrative measurements demonstrate a calculation, not a suggested body size, composition, or weight.
Worked female example: include the hip measurement
For a hypothetical adult aged 32, use height 64 in, neck 12.5 in, natural waist 29 in, hip 39 in, and weight 62 kg. The circumference combination is 29 + 39 − 12.5 = 55.5 in. Substituting into the female equation gives approximately 29.8554517%, displayed as 29.9%. The corresponding metric lengths are 162.56 cm, 31.75 cm, 73.66 cm, and 99.06 cm. Omitting hip leaves a required input missing rather than creating a simpler version of the equation.
At 62 kg, estimated fat mass is approximately 18.5104 kg and fat-free mass is approximately 43.4896 kg, displayed as 18.5 kg and 43.5 kg. Fat-free mass includes water, bone, organs, muscle, and other nonfat tissues; it is not a direct muscle measurement. Entered weight changes these two mass estimates but not the percentage. Age checks the tool's adult scope and is not an age coefficient in either historical equation.
Small measurement differences can move the predicted percentage
Keep the male example's height and neck fixed but change abdomen from 34 to 35 in. The circumference difference becomes 20 in, producing approximately 19.4290781% instead of 17.5130854%. That is an increase of about 1.916 percentage points, not a 1.916% relative increase. This deliberately large one-inch comparison illustrates sensitivity; it is not a typical measurement error estimate, an expected daily change, or a correction factor you should apply to a real result.
A changed prediction does not by itself establish how much tissue changed. Tape placement, posture, tension, rounding, and input mistakes can alter the calculation. Keep units and measurement sites consistent when comparing existing records, and do not select whichever reading gives the preferred answer. The tool has no saved profile, trend diagnosis, or individualized uncertainty interval. Reporting more decimal places would make the arithmetic look more detailed without improving the underlying measurement or model.
Why BMI, a body-fat scale, and DXA can disagree
BMI divides mass by squared height; it is not a body-fat percentage. The circumference model predicts a percentage from selected lengths. Bioelectrical impedance devices and DXA use other measurements and assumptions. Their outputs should therefore not be treated as identical quantities obtained by interchangeable instruments. The BMI guide explains its separate screening role. A discrepancy between methods is not automatically proof that one device is broken or that a calculator has used the wrong units.
Potter and colleagues' 2022 study compared circumference predictions, bioelectrical impedance, and DXA in 430 male and 179 female physically fit Marines aged 18–57. It found group-dependent differences and concluded that circumference predictions did not provide research-grade quantitative body composition data. That specific population cannot establish a universal personal error range for all adults. Do not subtract a study's average bias from your result as an individualized correction or assume that every consumer scale reproduces the study's equipment.
Use the tool without turning the estimate into a prescription
Choose metric or US mode, select the historical equation term, enter the required measurements and body weight, and check the labeled outputs. US height is total decimal inches, so 5 ft 10 in becomes 70 in rather than 5.10. Switching units converts existing entries. Blank required measurements, nonfinite numbers, invalid logarithm combinations, or predictions outside the tool's accepted percentage range produce an error rather than a clipped, plausible-looking answer.
The page does not implement current military pass/fail rules, service-specific rounding, calorie prescriptions, or a weight-loss plan. Historical source documents identify the equation, not today's occupational requirements. A software age or measurement boundary also does not prove that the model is appropriate for everyone inside that range. If you need a clinical assessment, have symptoms or relevant conditions, or find body-composition tracking distressing, seek qualified individual support rather than treating a population prediction as a health verdict.