FFMI (Fat-Free Mass Index)

FFMI: what is it?

BMI's best-known failure is that it counts a kilo of muscle and a kilo of fat identically, so lean, muscular clients land in the overweight band. FFMI is the obvious fix: strip the fat mass out first, then run the same height normalisation. What comes back is a number describing how much lean tissue someone carries for their height. It answers a question coaches are asked constantly, which is why a client with a "bad" BMI and visible abs is not, in fact, overweight. It also carries a serious catch, which is that it is only ever as good as the body fat estimate you fed it.

The formula, and the normalised version

Start by converting body fat percentage to fat-free mass: fat-free mass equals body weight multiplied by one minus the body fat fraction. Then divide by height in metres squared. So a client at 82 kg and 15 percent body fat carries roughly 69.7 kg of fat-free mass; at 1.80 m that is 69.7 divided by 3.24, giving an FFMI of about 21.5.

Because a height-squared index tends to score shorter people low and taller people high, a normalised version is often used to compare across heights. It adds 6.1 multiplied by 1.8 minus height in metres, which lifts shorter clients and pulls taller ones down. For a 1.80 m client the adjustment is zero; for someone at 1.70 m it adds about 0.6, and for someone at 1.90 m it subtracts about 0.6.

Typical values

The orientation figures below are widely used in the fitness world. Treat them as rough bands rather than precise cut-offs, because every one of them depends on how the underlying body fat was estimated.

The "natural limit of 25" claim

You will see it stated confidently that an FFMI above 25 is impossible without drugs. That figure comes from a small sample of bodybuilders and has been repeated so often that it now circulates as settled fact. It is not. The sample was small, the body fat estimates behind it carried the usual field-method error, and the argument says nothing about the extremes of the distribution or about how the number behaves in very tall or very short people.

Treat it as a rough population observation, not a test. Accusing a client of anything on the basis of an index derived from a caliper reading is not a defensible position for a coach to take.

What it is actually good for

FFMI earns its place in exactly one conversation: separating muscle from fat inside a high BMI. A client at BMI 28 with an FFMI of 23 and a low body fat estimate is muscular. A client at BMI 28 with an FFMI of 18 is not. That distinction changes how you talk to them and what you prioritise, and neither BMI nor scale weight can make it.

Its limitation is unavoidable and worth saying out loud: every gram of error in the body fat estimate flows straight through into the FFMI. If you measured body fat with bioimpedance on a dehydrated client, the FFMI is wrong by exactly as much. Which means the same rules apply as for any body composition figure. Same method, same tester, same conditions, and read the trend rather than the decimal.