Ffmi Calculator
A fat-free mass index (FFMI) calculator measures how much lean mass you carry for your height, independent of body fat. Enter your weight, height, body fat percentage and sex to get your raw FFMI, the height-normalized FFMI, your fat mass index (FMI), and an honest reading of your score against the research, including the well-known 25.0 reference point described by Kouri and colleagues in 1995. The score is only as good as the body fat figure behind it, and this page shows exactly how much a body fat error moves the result.
Quick Answer
A man weighing 80 kg at 178 cm with 15% body fat carries 68.0 kg of fat-free mass. His raw FFMI is 21.46, his height-normalized FFMI is 21.59 (band: Above average), his fat mass index is 3.79, and the cross-check holds: 21.46 + 3.79 = a BMI of 25.25, split into lean and fat parts.
Key takeaways
- FFMI answers the question BMI cannot: how much of your weight is lean tissue, scaled to your height.
- Compare people on the normalized score. The bands and the 25.0 reference point were published on normalized values.
- The 25.0 figure is a research finding from one 1995 sample of male athletes, not a verdict on any individual.
- Your body fat input is the weak link. A 5-point underestimate moves the worked example’s score by about 1.3 points.
- Quick Answer
- How to use the FFMI calculator
- The FFMI formula, step by step
- Worked example: 80 kg, 178 cm, 15% body fat
- Reading the bands: men and women
- The 25.0 reference point, stated honestly
- The 6.3 factor, taken from the paper itself
- Your body fat input runs the result
- What competing calculators miss
- Frequently asked questions
How to use the FFMI calculator
Enter four things and pick your sex, then click Calculate. The results stay hidden until you do.
- Weight and height, in either unit system. Pounds convert at 0.453592 kg and inches at 2.54 cm, exactly as the original tool converted them.
- Body fat percentage. Use the best measurement you have, and use the same method every time you re-test, or the change you see may be method noise rather than muscle.
- Read the normalized score first. It is the headline number, the one the band and the 25.0 discussion refer to. The raw FFMI, fat-free mass, fat mass, FMI and the BMI cross-check sit underneath it.
The FFMI formula, step by step
FFMI is built in three steps: strip the fat off the weight, scale what remains to height, then correct for height itself.
FFMI = Fat-free mass (kg) ÷ Height (m)2
Normalized FFMI = FFMI + 6.3 × (1.80 − Height in m)
The height correction exists because raw FFMI drifts with stature: at the same build, a taller person’s raw score and a shorter person’s raw score are not measuring the same thing on the same scale. Normalizing to 1.80 m, the reference height Kouri and colleagues used, removes most of that drift. The factor in the formula above is 6.3 because that is the correction the published abstract of the 1995 paper states: according to Kouri et al. on PubMed, they “added a slight correction of 6.3 x (1.80 m – height) to normalize these values to the height of a 1.8-m man.” A 6.1 variant circulates in some later reviews and calculators; a full section below quantifies what that variant would change, which is a few hundredths of a point at ordinary heights.
Worked example: 80 kg, 178 cm, 15% body fat
The pre-filled values are the worked example, so the tool reproduces every figure in this section. Fat-free mass is 80 × (1 − 0.15) = 68.0 kg and fat mass is 12.0 kg. FFMI is 68.0 ÷ 3.1684 = 21.46. The normalization adds 6.3 × (1.80 − 1.78) = 0.126, giving 21.59, which lands in the men’s Above average band. FMI is 12.0 ÷ 3.1684 = 3.79, and 21.46 + 3.79 = 25.25, the man’s BMI, which is the built-in arithmetic check that the lean and fat parts were split correctly.
| Check run | Raw FFMI | Normalized | Band |
|---|---|---|---|
| Man, 80 kg, 178 cm, 15% (worked example) | 21.46 | 21.59 | Above average |
| Woman, 65 kg, 165 cm, 22% | 18.62 | 19.57 | Excellent |
| Man, 180 lb, 70 in, 15% (imperial entry) | 21.95 | 22.09 | Excellent |
| Man, 100 kg, 200 cm, 10% (tall: normalization subtracts) | 22.50 | 21.24 | Above average |
The last row is the one that teaches the normalization: at 2.00 m the correction is 6.3 × (1.80 − 2.00) = −1.26, pulling a flattering raw score down to the 1.80 m scale. Below 1.80 m the correction runs the other way and lifts the score.
Reading the bands: men and women
The badge under your result reads your normalized score against the band table for your sex. These are the exact bands the calculator applies:
| Normalized FFMI, men | Reading | Normalized FFMI, women | Reading |
|---|---|---|---|
| Under 18 | Below average | Under 14 | Below average |
| 18 to under 20 | Average | 14 to under 16 | Average |
| 20 to under 22 | Above average | 16 to under 18 | Above average |
| 22 to under 24 | Excellent | 18 to under 20 | Excellent |
| 24 to under 25 | Near the natural ceiling | 20 to under 21 | Near the estimated natural ceiling |
| 25 and above | Above the drug-free observation | 21 and above | Above the estimated natural ceiling |
Keep the bands in perspective with general-population data. In the Schutz, Kyle and Pichard reference study of 5,635 healthy adults, the median FFMI for young adults was 18.9 in men and 15.4 in women, squarely inside this tool’s Average bands. The upper bands describe trained lifters, not the general public, and the women’s ceiling is labelled an estimate because the underlying ceiling research sampled men.
The 25.0 reference point, stated honestly
The famous natural limit comes from one paper: Kouri, Pope, Katz and Oliva (1995), Fat-free mass index in users and nonusers of anabolic-androgenic steroids. They measured 157 male athletes, 83 steroid users and 74 nonusers, and found the nonusers’ normalized FFMI values extended up to a well-defined limit of 25.0, while many users exceeded it, some past 30. A comparison sample of 20 Mr. America winners from the pre-steroid era averaged 25.4. The authors proposed FFMI as a useful initial screening measure, and they called their own findings preliminary.
One technical point matters for comparing your score to that limit: the 25.0 figure was computed on the 6.3 normalization, because that is the correction the paper’s abstract states it applied before reporting the normalized values. This tool uses the same 6.3 factor, so its normalized scores sit on exactly the paper’s scale and no conversion is needed before reading a score against 25.0.
Three caveats keep that finding honest. It was one study of 74 drug-free men, measured with skinfolds, not a universal law of physiology. The pre-steroid champions averaging 25.4 show the line was never a hard wall even in the paper’s own data. And a score above 25 says at least as much about the body fat estimate as about the body: the calculator’s own note under a high result tells you to re-check the body fat figure first, which is the right order of operations. Read 25.0 as the point where a score becomes unusual and worth double-checking, never as evidence about a person.
The 6.3 factor, taken from the paper itself
This tool normalizes with 6.3, the factor stated in the published abstract of Kouri et al. (1995): the authors write that they added “a slight correction of 6.3 x (1.80 m – height)” to normalize their values. A 6.1 variant also circulates, printed in some peer-reviewed reviews of the method and used by some other calculators, but it is not the figure in the paper’s abstract, so it is not the figure this tool uses. Here is what the 6.3 correction itself does at real heights, for a raw FFMI held constant, and what picking the 6.1 variant instead would change:
| Height | Adjustment at 6.3 (this tool) | What the 6.1 variant would change |
|---|---|---|
| 1.50 m | +1.89 | 0.06 |
| 1.65 m | +0.95 | 0.03 |
| 1.80 m | 0.00 | 0.00 |
| 1.95 m | −0.95 | 0.03 |
| 2.10 m | −1.89 | 0.06 |
The practical verdict: the 6.1 variant would move a score by at most a few hundredths inside normal height ranges, far less than one point of body fat error. It cannot flip a band except for a score sitting exactly on a boundary, and it never explains a full-point disagreement between two calculators. That disagreement is almost always the body fat input, which the next section prices out. What the 6.3 choice does settle is provenance: the score, the bands and the 25.0 reference point are all read on the scale the original paper published.
Your body fat input runs the result
FFMI inherits every error in the body fat figure, magnified. Hold the worked example’s weight and height fixed and vary only the body fat entry:
| Body fat entered (80 kg, 178 cm, man) | Normalized FFMI | Band |
|---|---|---|
| 10% (5 points too low) | 22.85 | Excellent |
| 15% (the worked example) | 21.59 | Above average |
| 20% (5 points too high) | 20.33 | Above average |
Each body fat point is worth about a quarter of an FFMI point for this build, so a 5-point underestimate flatters the score by roughly 1.3 points and can carry it across a band boundary. Method choice is where those points hide: in a PLOS ONE validation against DXA, bioelectrical impedance underestimated body fat by a mean of 4.6 points in men and 5.1 in women at baseline, while studies in trained adults find multi-frequency impedance tracks DXA change scores much better than its one-off readings. DXA and careful skinfolds by a practiced measurer sit at the reliable end. Whatever you use, keep the method fixed between tests. For planning what a higher score would require in practice, the bulking calories calculator and the clean bulk calculator cover the surplus side of gaining lean mass, and the body surface area calculator gives the other classic height-and-weight body index for comparison.
This calculator is for education and training tracking only and is not medical advice. Body composition interpretation depends on individual health context; discuss yours with a clinician.
What competing calculators miss
The FFMI web is split on its own formula and loose with its scariest label. The patterns that matter:
- The uncited-factor camp: calculators on both sides of the 6.1 and 6.3 split print a factor without citing where it comes from. The paper’s abstract states 6.3, and a reader comparing two calculators deserves to be told which figure each one uses and why.
- Verdict-first charts (GCalculate and similar): labels like “Suspicious” and “Steroid confirmed” attached to score ranges. No study confirms anything about an individual from an index value, and the original authors proposed a screening prompt, not a diagnosis.
- Raw-score-only tools: skip normalization entirely, so tall and short lifters get compared on a tilted scale and the published bands, which are normalized, get applied to the wrong number.
- The gap this page fills: both scores with the normalization shown, the exact band tables the code applies, the 25.0 finding presented with its sample size and its authors’ own “preliminary” label, the 6.3 factor quoted from the paper’s abstract with the 6.1 variant’s difference priced at real heights, and a sensitivity table showing what a body fat error does before you trust any of it.
Frequently asked questions
For men, a normalized FFMI of 20 to 22 is a solid trained level and 22 to 24 is excellent. For women, 16 to 18 is a solid trained level and 18 to 20 is excellent. Population medians sit lower, around 18.9 for young men and 15.4 for young women in the Schutz reference data, because most people do not train for muscle.
No. Kouri and colleagues proposed FFMI as an initial screening measure, and their drug-free sample topped out at a normalized 25.0, but a score above that line is a prompt to re-check the body fat measurement, not proof about any person. Body fat underestimates inflate FFMI quickly, and the authors themselves called their findings preliminary.
Raw FFMI drifts with height even at identical muscularity, so scores for a 1.65 m lifter and a 1.95 m lifter are not directly comparable. The normalization adds 6.3 times the difference between 1.80 m and your height, putting everyone on the 1.80 m scale the research bands were built on.
Use the normalized score for the research bands and the 25.0 reference point, because those were published on normalized values. The raw score is still useful for tracking your own change over time at a fixed height, where the adjustment is a constant and cancels out.
The fat mass index is fat mass divided by height squared, the mirror image of FFMI. The two add up exactly to BMI, so together they split one blurred number into its lean and fat parts: two people with a BMI of 25 can carry very different amounts of muscle and fat.
As accurate as you can manage, because every point of body fat error moves the score. For the worked example above, reading 10 percent instead of a true 15 percent lifts the normalized FFMI from 21.59 to 22.85. DXA and careful skinfold measurement sit at the reliable end; single-frequency bathroom scales have underestimated body fat by several points against DXA in validation studies.
The research is thinner for women because the 1995 study sampled men only. A normalized FFMI around 21 is commonly cited as the approximate upper range for drug-free women, and this calculator labels women’s bands as estimates for that reason. Treat any women’s ceiling as a rough guide, not a measured limit.
Two reasons account for nearly every mismatch. Some calculators normalize with a 6.1 factor where this one uses the 6.3 factor stated in the Kouri et al. abstract, which moves scores by a few hundredths at most heights. The bigger cause is the body fat figure: different measurement methods, or the same method on a different day, can shift body fat by several points and FFMI by a full point or more.
Add This Calculator to Your Website
Embed this calculator on your own site with the code below. Pick a size to preview it live, then copy the code.
Powered by Calculator Zilo
References
- Kouri, E.M., Pope, H.G., Katz, D.L. and Oliva, P. Fat-free mass index in users and nonusers of anabolic-androgenic steroids (Clinical Journal of Sport Medicine, 1995; PubMed record). Accessed October 21, 2026.
- Raschka, C. Side effect profiles and doping stigmata: review printing the FFMI normalization with the 6.1 factor (Papers on Anthropology, University of Tartu). Accessed October 21, 2026.
- Schutz, Y., Kyle, U.U.G. and Pichard, C. Fat-free mass index and fat mass index percentiles in Caucasians aged 18 to 98 years (International Journal of Obesity, 2002). Accessed October 21, 2026.
- Kyle, U.G., Schutz, Y., Dupertuis, Y.M. and Pichard, C. Body composition interpretation: contributions of the fat-free mass index and the body fat mass index (Nutrition, 2003). Accessed October 21, 2026.
- Relationship of fat mass index and fat free mass index with BMI and association with function, cognition and sarcopenia in pre-frail older adults. PMC full text (PMC8741276). Accessed October 21, 2026.
- Increased odds of having the metabolic syndrome with greater fat-free mass: NHANES fat-free mass index percentiles. Journal of Cachexia, Sarcopenia and Muscle, Wiley Online Library. Accessed October 21, 2026.
- Li, C. et al. Percentage of body fat assessment using bioelectrical impedance analysis and dual-energy X-ray absorptiometry (PLOS ONE, PMC3613423). Accessed October 21, 2026.
- Comparison of DXA versus a multi-frequency bioelectrical impedance device for body composition assessment after a 4-week hypoenergetic diet. PMC full text (PMC7739224). Accessed October 21, 2026.
- Comparison of bioelectrical impedance analysis and dual energy X-ray absorptiometry with a three-compartment model. PMC full text (PMC7177846). Accessed October 21, 2026.
- Coin, A. et al. Fat-free mass and fat mass reference values by dual-energy X-ray absorptiometry in a 20 to 80 year old Italian population (Clinical Nutrition). Accessed October 21, 2026.
- Kanayama, G., Hudson, J.I. and Pope, H.G. Risk factors for illicit anabolic-androgenic steroid use in male weightlifters (Biological Psychiatry, PMC3218214). Accessed October 21, 2026.