Learn

FFMI: what fat-free mass index actually tells you

BMI treats a 90 kg powerlifter and a 90 kg office worker of the same height as the same person, because weight is all it knows. Fat-free mass index is the number that tells them apart.

Last reviewed

How it is worked out

FFMI is your fat-free mass divided by your height squared — the same shape as BMI, with the fat taken out first:

fat-free mass (kg) = weight × (1 − body fat %)
FFMI = fat-free mass ÷ height (m)²

Someone of 80 kg and 1.80 m at 20% body fat is carrying 64 kg of fat-free mass, so their FFMI is 64 ÷ 3.24, or 19.8. The same person at the same weight with 15% body fat has an FFMI of 21.0 — same scales, same height, meaningfully different body.

The height adjustment

Dividing by height squared does not fully cancel height out, and the raw figure quietly favours shorter people. The 1995 paper that popularised FFMI applied “a slight correction of 6.3 × (1.80 m − height)”, normalising everyone to a 1.80 m reference. At 1.70 m that adds 0.63 to the raw number; at 1.90 m it takes 0.63 away.

Two things worth knowing. Most calculators online use 6.1 rather than 6.3, which is not the constant the original paper used — the difference is small but it is not nothing, and almost nobody says which they are using. And an adjusted FFMI is not comparable with an unadjusted one, so a figure from one calculator cannot be read against a figure from another.

What counts as a high or low figure

There is no clinical classification for FFMI in the way there is for BMI. What exists is a set of conventional bands, and they are descriptive labels rather than diagnoses. Tare uses these, for men:

  • Below 17 Below average
  • 17 – 19 Average
  • 19 – 21 Above average
  • 21 – 23 Excellent
  • 23 – 26 Superior
  • 26 and above Exceptional

Reference figures for women are considerably thinner on the ground, and are lower across the board — the research FFMI grew out of studied men almost exclusively, and that gap has never really been filled.

Where “25 is the natural limit” came from

This is the claim FFMI is famous for, and it comes from a single paper. In 1995 Kouri, Pope, Katz and Oliva measured 157 male athletes — 83 who had used anabolic steroids and 74 who had not — and reported that the normalised FFMI of the non-users “extended up to a well-defined limit of 25.0”, while users routinely passed it and some exceeded 30. They proposed FFMI as a first-pass screen for steroid use.

That is a reasonable finding, and it is not what the number has since been turned into. A screening threshold and a limit of human potential are different claims, and the second one does not survive much scrutiny:

  • 74 people is a small sample for an outer bound. No trait’s extreme — height, sprint speed, anything — would be found by measuring 74 people who happened to be in a Boston or Los Angeles gym.
  • The measurement error is bigger than the margin. Body fat was taken by skinfold, which carries a standard error of around 3.4%. Nine of the non-users landed between 24.0 and 24.9, and at least one measured 25.4 — above the limit the paper is remembered for.
  • The paper’s own historical comparison exceeds it repeatedly. Its analysis of Mr America winners from 1939–1959, before steroids were available, found a mean FFMI of 25.4 ± 1.5, with 13 of 20 above 25 and a peak of 28.0. Their body fat was estimated by eye from magazine photographs.
  • Later data does not agree. A 2024 study of 111 NCAA Division III American football players found a mean FFMI of 23.6 ± 2.0 and a range of 18.1 to 27.7, with 21% of the squad above 25. Its authors concluded plainly that the limit of 25 “underestimates that observed in American collegiate football players”.

So 25 is a useful signpost — a figure very few people reach, and reaching it says something. It is not a wall, and an FFMI above it is not evidence about anybody.

The weak link is your body-fat figure

FFMI inherits every bit of error in the body-fat estimate feeding it, and most people’s body-fat figure is an estimate from a tape measure or a formula rather than a measurement. Take the same 80 kg person at 1.80 m:

  • 15% body fat FFMI 21.0
  • 20% body fat FFMI 19.8
  • 25% body fat FFMI 18.5

A five-point error in the body-fat figure moves FFMI by more than a point — wider than the gap between two of the bands above, without a gram of anything changing. Which is the real lesson: the trustworthy thing is your own FFMI over time, measured the same way each time. The absolute number, compared against someone else’s, is carrying more uncertainty than it looks like it is.

What FFMI cannot tell you

Fat-free mass is not muscle mass. It includes bone, organs, connective tissue and the water you are carrying that day — so a dehydrated reading and a well-hydrated one are different numbers about the same body. FFMI also says nothing about where mass sits, and nothing at all about strength, fitness or health. It is one number about composition, and it is at its most useful sitting next to others rather than on its own.

Sources

  • Kouri EM, Pope HG Jr, Katz DL, Oliva P. “Fat-free mass index in users and nonusers of anabolic-androgenic steroids.” Clinical Journal of Sport Medicine, 1995 — the origin of the normalised FFMI and of the 25.0 figure.
  • Fields JB, Jones MT, Kuhlman NM, Magee MK, Feit A, Jagim AR. “Fat-Free Mass Index in a Large Sample of Collegiate American Football Athletes.” International Journal of Exercise Science, 2024;17(4):129–139.
  • Nuckols G. “What Everyone Gets Wrong About FFMI and the ‘Natty Limit’”, 2016 — a reappraisal of the Kouri data, including the raw values and the Mr America comparison.

This page is reference information about how a measurement is taken and what the research says about it. It is not medical advice, and no figure on it is a diagnosis. Talk to a doctor, nurse or registered dietitian about your own health.