What FFMI measures, normalized FFMI vs FFMI, and its limits

FFMI needs a body-fat estimate before it can be calculated

FFMI and fat mass index, or FMI, divide the two estimated parts of body weight by the same height-squared value. Before rounding, FFMI plus FMI equals BMI. The split depends on the body-fat percentage you enter, so a different tape, scan, or prediction method can change FFMI even when weight and height stay the same.

Fat-free mass includes more than skeletal muscle. It includes water, bone, organs, and other non-fat tissue. A high or low FFMI result does not show how much skeletal muscle you have. It also does not show strength, training response, health, or a change in muscle tissue.

Body Numeral calculates FFMI from weight, height, and a body-fat percentage that you provide. Keep the name of the body-fat method beside the result. Compare FFMI only when the body-fat method, height unit, and height-adjustment convention are the same.

Normalized FFMI is FFMI with a height correction added. Unadjusted FFMI, also called plain or raw FFMI, is fat-free mass in kg divided by height in metres squared. Normalized FFMI adds 6.3 × (1.80 − height in m) to that number. The 1995 study abstract says it added this correction to normalize the values to the height of a 1.8 m man. Height-adjusted FFMI and adjusted FFMI are other names for the same thing.

The correction raises the number for people shorter than 1.80 m and lowers it for people taller than 1.80 m. At exactly 1.80 m the two numbers are the same. With 6.3, each centimetre away from 1.80 m moves the result by 0.063. For a hypothetical person at 170 cm with an FFMI of 22: 22 + 6.3 × (1.80 − 1.70) = 22 + 0.63 = 22.63, shown as 22.6.

Hypothetical people who all have an unadjusted FFMI of 22 kg/m²

The correction does not change fat-free mass, body fat or weight. It changes only the index. Many calculators use 6.1 instead of 6.3. Between 150 cm and 210 cm the two coefficients give results no more than 0.06 apart before rounding (0.1 after rounding to one decimal). The choice of coefficient changes the result much less than the choice between unadjusted and normalized FFMI.

Compare like with like. The percentile tables further down use unadjusted FFMI. The 25 from the 1995 study is a normalized value.

Height-adjusted FFMI sets every result against a height of 1.80 m, so the same adjusted value means a different amount of fat-free mass at each height. With the 6.3 adjustment from the 1995 study, an adjusted FFMI of 25 takes 68.1 kg (150 lb) of fat-free mass at 5 ft 6 in and 90.1 kg (199 lb) at 6 ft 2 in. The table runs the calculator's own equation for heights from 5 ft 0 in to 6 ft 6 in.

Fat-free mass at an adjusted FFMI of 21 and 25 (6.3 adjustment)

To turn fat-free mass into body weight, divide by (1 − body fat / 100). An adjusted FFMI of 25 equals an unadjusted 25 only at 1.80 m. At 5 ft 6 in it equals an unadjusted 24.2, and at 6 ft 4 in an unadjusted 25.8. With the 6.1 adjustment some calculators use, each value changes by no more than 0.2 kg or 1 lb.

The FFMI calculator shows how many kg of fat-free mass your result is below adjusted FFMI 21 and 25. It shows 0 at or above each value. In the 1995 study of male athletes, 25 was the upper end seen in those who had not used steroids. The 21 does not come from that study. It is the top of the calculator's own female comparison bands, which are extrapolated. Neither number is a target, a personal limit or a test for drug use.

Reference values depend on how fat-free mass was measured and who was studied. A 2002 cross-sectional study reported FFMI and FMI percentiles in 5,635 apparently healthy Caucasian adults in Switzerland, ages 24 to 98, using bioelectrical impedance with formulas cross-validated against dual-energy X-ray absorptiometry. Those percentiles do not become universal targets or diagnostic cutoffs.

A large Korean national sample was measured with bioelectrical impedance. For ages 20 to 59, its age-group medians ranged from 18.6 to 19.0 kg/m² for men and from 14.7 to 15.5 kg/m² for women. The Swiss study above reported close medians for adults aged 18 to 34: 18.9 kg/m² for men and 15.4 kg/m² for women. The tables below show how values spread by age. They describe the people measured in each study. A percentile is not a target, and it does not show health or fitness.

Men: FFMI percentiles by age, Korea 2022 to 2023 (kg/m²)

Women: FFMI percentiles by age, Korea 2022 to 2023 (kg/m²)

Predicted average FFMI by age, adults at one New York center measured with DXA (kg/m²)

Compare these tables with unadjusted FFMI. The tables do not use a height adjustment. The Body Numeral FFMI calculator shows the normalized figure first, with the 6.3 or 6.1 coefficient you select, and the unadjusted FFMI in the rows below it. The calculator's comparison labels are its own conventions, not values from these studies.

Here is an example for men aged 30 to 39 in the Korean sample. The median is 19.0, the 75th percentile is 20.3, and the 95th percentile is 22.5. An unadjusted FFMI of 20 falls between the median and the 75th percentile. A value of 22 falls between the 75th and 95th percentiles. A value of 25 is above the 95th percentile in every adult age group in that sample. For women aged 30 to 39, the median is 15.2, the 75th percentile is 16.2, and the 95th percentile is 18.6.

The 25 often quoted online comes from the preliminary 1995 study of 157 male athletes. It is height-adjusted to 1.80 m and describes that group of athletes, so it is not a point on these population tables.

The two samples differ in country, ancestry, and measurement method, so their numbers are not interchangeable. If you can, compare your result with a sample measured by a method like yours. Then keep that same method for later checks, so a change in method does not look like a change in your body.

Use a qualified professional when a body-composition estimate affects nutrition, training, or medical decisions. A calculator result can organize a method and an input set. It cannot replace a clinical assessment.

1. Save the body-fat percentage and the method that produced it.
2. Save body weight, height, units, date, and the height-adjustment convention.
3. Compare results only after checking that the same calculation and body-fat method were used.
4. Keep a scan, tape estimate, and prediction equation in separate series.

Open the FFMI calculator to inspect the arithmetic. Use the muscle-mass guide when you need to separate a skeletal-muscle estimate from fat-free mass.

Take a hypothetical 80 kg person at 1.80 m with an estimated body-fat percentage of 20%. The estimated fat mass is 80 × 0.20, or 16 kg. Fat-free mass is 80 − 16, or 64 kg. Height squared is 1.80 × 1.80, or 3.24 m². FFMI is 64 ÷ 3.24, about 19.75 kg/m². This is an arithmetic example, not a measured amount of muscle. If the body-fat estimate came from another method, the 64 kg starting value could change. Keep the percentage source with the FFMI. Without it, another reader cannot reproduce or properly compare the index.

The denominator in FFMI is height in metres squared, not centimetres squared. At 1.80 m, the denominator is 3.24 m². Squaring 180 cm gives 32,400 cm², which is mathematically equivalent only after the unit conversion; entering 180 as though it were metres creates a meaningless displayed index. Body Numeral handles its supported unit inputs, but a copied calculation should show the height conversion before division. Keep the unrounded height and fat-free-mass values while checking the result. The index has units of kg/m² even though people often quote only its number. That unit distinguishes it from fat-free mass in kilograms and from a body-fat percentage.

Fat mass index, or FMI, divides estimated fat mass by the same height squared. In the 80 kg example, FMI is 16 ÷ 3.24, about 4.94 kg/m². Before rounding, FFMI plus FMI is 80 ÷ 3.24, about 24.69 kg/m², which is BMI. This identity is a useful arithmetic check. It does not validate the entered body-fat percentage or prove that the fat-free portion is skeletal muscle. If your displayed sum differs slightly from BMI, check whether the calculator rounded each component before adding it. Retain the unrounded values when you audit a result.

At the same 80 kg and 1.80 m, changing only the estimated body-fat percentage from 20% to 22% lowers calculated fat-free mass from 64 to 62.4 kg. FFMI then becomes about 19.26 kg/m². The index changed by about 0.49 because the body-fat input changed by two percentage points. The calculation does not show that 1.6 kg of muscle was lost. It shows the result’s dependence on one estimate. A tape equation, bioimpedance device, or scan can yield different percentages; the FFMI values derived from them should remain labeled by method rather than averaged.

Save whether a result is unadjusted or normalized FFMI, and which coefficient produced it. The normalized FFMI vs FFMI table earlier on this page shows how far the correction moves the number at each height. Do not compare a normalized value from one tool with an unadjusted value from another as though they used one formula.

A published percentile describes the people and method used to create it. The cited Swiss adult study used a bioimpedance-based method, and the cited 1995 athlete sample included selected male athletes. Neither sample gives a universal health target or a personal ceiling. A value near a number discussed in an athlete paper cannot establish drug use. A value below a population percentile cannot diagnose low muscle. If a clinician needs to assess muscle, the appropriate test and reference depend on the clinical question. Body Numeral’s FFMI is best read as a transparent calculation from weight, height, and a named body-fat estimate.