Body fat percentage chart by age and sex

Body fat percentage is the share of your total body mass that is adipose tissue rather than muscle, bone, organ, water and everything else. It is a two-compartment description of a body: fat mass and fat-free mass, expressed as a ratio. That is genuinely more informative than weight, because it separates the two kinds of kilogram that a scale adds together.

The awkward part is that nothing in routine use measures it. Fat cannot be weighed separately while you are still using it, so every method infers the split from something else. Hydrostatic weighing and air displacement infer it from whole-body density. DEXA infers it from how two X-ray energies are attenuated by tissues of different composition. Bioelectrical impedance infers it from how a small current travels through body water. Skinfold calipers infer it from the thickness of subcutaneous fat at a handful of sites. Tape-measure formulas such as the US Navy method infer it from circumferences.

Each of those steps carries an assumption (that lean tissue has a fixed density, that hydration is normal, that subcutaneous fat predicts the internal amount), and each assumption fails in some people. That is why a figure reading 22% on one device can read 18% on another in the same hour. The practical error band against a DEXA reference is about three to five percentage points for the field methods, and DEXA itself varies between manufacturers.

None of which makes the number useless. It makes it a range. A body fat estimate of 22% is honestly saying 'probably somewhere between 18 and 26', which is still enough to tell you which part of the chart you are in, and (measured the same way each time) enough to track a real change over months.

The American Council on Exercise categories are the ones most people have seen. They are sex-specific, deliberately simple, and framed around training status rather than disease risk: an essential-fat floor, an athlete band, a fitness band, an average band, and a threshold above which the term obese is applied.

Their strength is that they are easy to act on and they acknowledge the large, genuine difference between male and female bodies. Their weakness is that they apply the same numbers to a 22-year-old and a 68-year-old, which the evidence does not support. Read them as descriptions of training status in an adult population, not as clinical thresholds.

The second system takes a different route. Gallagher and colleagues measured body fat in 1,626 adults across three sites (Cambridge in the United Kingdom, Columbia in New York and Jikei in Japan) using four-compartment models and DEXA, then built regression equations predicting body fat from the inverse of BMI, sex, age and ethnicity. They then solved those equations at the three BMI values that already define underweight, overweight and obesity: 18.5, 25 and 30.

The result is a set of body fat percentages that correspond to the weight guidelines already in use. The table below is Table 4 of that paper, the combined figures for white and African American subjects. Read a row across: anything below the first column is the body fat equivalent of being underweight, the gap between the first and second columns is the healthy range, the gap between the second and third is the overweight equivalent, and the third column onwards is the obesity equivalent.

Two honesty notes the paper itself insists on. First, the authors describe these as provisional ranges and a working approach, not as settled guidelines; their stated aim was to show how such guidelines could be built. Second, the standard error of the estimate behind this table is 4.98 percentage points, and the sample contained few people at very low BMI, so the bottom of the scale has wide confidence intervals and should be applied cautiously.

The same paper found that a single set of universal ranges could not be built, because at the same BMI the Asian cohort carried a different amount of fat. It therefore published separate equations and a separate table, reproduced below. The pattern mirrors the one on our BMI chart page: at a given BMI, body fat runs higher, so the percentage that corresponds to any given weight category is higher too.

This is the same argument that produced the WHO Asia-Pacific BMI cut-offs, arriving from the opposite direction. If you are reading a body fat number against a BMI-derived category, the population the equation was fitted on matters.

Put a 55-year-old woman at 30% body fat against both charts. ACE places her in the average band, two points below its obese threshold of 32%. Gallagher places her comfortably inside the healthy range for her age, whose overweight equivalent does not begin until 34%. Neither chart is wrong. They were built to answer different questions.

ACE's categories describe fitness and training status, drawn from the exercise-science literature, and their reference point is an active adult. Gallagher's ranges describe the body fat that corresponds to weight categories already tied to population health outcomes, and their reference point is epidemiology. A number that looks unremarkable through one lens can look raised through the other, and the sensible response is to know which lens you are using rather than to pick the flattering one.

The practical way to read both: use ACE when the question is about training and fitness, use Gallagher when the question is about where a number sits relative to established health guidelines, and if the two disagree, treat that disagreement as information. It usually means the number is near a boundary, and boundaries are lines drawn on a continuum. Then do what neither chart can: add a waist measurement, because where the fat sits carries risk information that a single percentage cannot.

• ACE is sex-specific but age-blind; Gallagher is both sex- and age-specific.
• ACE's labels describe training status; Gallagher's are anchored to BMI-based health guidelines.
• Gallagher publishes population-specific tables; ACE publishes one table for everyone.
• Neither system was validated against a field method; both assume a reference-quality measurement.

Look down the Gallagher columns rather than across them. In men, the body fat corresponding to BMI 25 rises from 20% in the 20–39 band to 22% at 40–59 and 25% at 60–79. In women it moves from 33% to 34% to 36%. The weight category has not changed; the body underneath it has.

The mechanism is well documented. From roughly the fourth decade, fat-free mass declines steadily (skeletal muscle most visibly, but bone mineral too) while fat mass is retained or increased, and fat also redistributes from subcutaneous stores toward the abdomen and around the organs. Two people can therefore hold the same weight and the same BMI thirty years apart and have materially different body composition. The Gallagher paper reports exactly this: after controlling for BMI, greater age was independently associated with a higher percentage of body fat.

This is why an age-blind chart misleads in both directions. It calls an older adult overfat at a composition that is entirely ordinary for their decade, and it can pass a younger adult at a composition that is genuinely raised for theirs. The paper is careful to note what remains unanswered: whether the extra fatness that comes with age at a constant BMI carries additional health risk is, in the authors' words, an important and as yet unanswered question. The range shifting upward describes what is typical, not what is optimal.

The gap between the male and female columns is not a scoring adjustment. It is a description of two differently constructed bodies, and it is present on every chart ever published, from ACE's categories to Gallagher's regressions.

Female bodies carry sex-specific adipose tissue that male bodies do not: deposits in breast tissue, around the pelvis and the hips, and in the intramuscular and intermuscular spaces. These stores are laid down under the influence of oestrogen from puberty onwards and are functional rather than surplus; they support reproductive capacity and serve as an energy reserve. Male bodies, under the influence of testosterone, carry more skeletal muscle for a given height and preferentially store fat viscerally and around the abdomen.

The practical consequence is that the entire female scale sits roughly ten percentage points higher than the male one, at every category, on every system. Under the ACE bands the female athlete range and the male average range overlap almost exactly; the same number describes completely different bodies. A woman at 22% body fat and a man at 22% body fat have very little in common physiologically, which is why a unisex body fat chart is not a simplification but an error.

Competitive athletes cluster below the general population, but the spread between sports is wider than the spread within most of them, and it runs in the direction the sport's demands would predict. Endurance sports where the body is carried against gravity or over distance (distance running, road cycling, triathlon, lightweight rowing) tend to show the lowest values, because mass that does not produce power is a cost. Aesthetic and weight-class sports sit low for a different reason: the sport itself judges or restricts body mass. Strength and power sports show mid-range percentages attached to very high absolute lean mass. Sports with a thermal or collision component, and several throwing and heavyweight categories, show values at or above the population average, and in those events that is an advantage rather than a deficiency.

We have deliberately not printed a table of sport-by-sport figures. The published numbers come from small cohorts, measured decades apart, by methods that disagree with one another by several percentage points, and they get copied between websites until the original sample is untraceable. A table like that would look authoritative and mean very little. If you want a defensible figure for a sport, the source is a current position stand or a governing body's own testing data, measured by a stated method.

The general caution is worth more than any table: an athletic body fat percentage is a by-product of training and a sport's demands, not a target to be pursued for its own sake. The IOC's 2023 consensus statement is explicit that many sports have entrenched cultures in which coaches and support staff exert pressure over body weight and composition, that this pressure is a documented route into low energy availability and disordered eating, and that body composition assessment should be undertaken only where it is genuinely justified and handled by people trained to do it safely.

Every chart on this page has a bottom, and the bottom is not an achievement marker. Essential fat is the lipid the body cannot function without: it is structural in cell membranes, nerve sheaths, bone marrow and the organs themselves, and in female bodies it includes sex-specific stores. It is not stored energy, and it cannot be reduced without a cost.

The values below are the essential-fat band from the same engine that drives the tables above, together with the point at which the athlete band begins, which in practice is the lowest level at which the published categories are still describing training rather than risk.

Sustained energy intake below what training and daily life demand produces a recognised clinical syndrome. The International Olympic Committee's 2023 consensus statement calls it Relative Energy Deficiency in Sport, or REDs, and describes it as a set of health and performance consequences that follow from low energy availability in both female and male athletes. Low body fat is not the cause (inadequate energy is), but a body fat percentage falling toward or below the essential-fat line is one of the more visible signs that energy availability has been low for a while.

The documented consequences are not cosmetic. The statement lists disruption of the hypothalamic–pituitary–gonadal axis, reduced testosterone in men and menstrual dysfunction up to functional hypothalamic amenorrhoea in women, reduced bone mineral density and increased bone stress injury, impaired immune function, altered metabolic and thyroid hormones, cardiovascular and gastrointestinal effects, and impaired iron and haematological status. In adolescents it can affect growth. Several of these, bone in particular, are slow to reverse.

None of this means a low number is automatically a problem. Some people are genuinely lean on adequate energy intake, and the statement is careful to distinguish adaptable from problematic low energy availability. What matters is the pattern around the number: whether periods have become irregular or stopped, whether injuries and illnesses are accumulating, whether recovery has stalled, whether performance is drifting downwards, whether food has started occupying a lot of mental space. Those signs are more informative than the percentage itself, and they are the reason this page does not encourage anyone toward a lower figure.

Body fat percentage is a ratio, and a ratio can move because either side moves. Someone who gains five kilograms of muscle and no fat sees their body fat percentage fall without losing a gram of fat. That is why the fat-free mass index is a useful companion: it expresses lean mass relative to height squared, the same way BMI treats total mass, and answers the question the fat number cannot: how much muscle is actually there.

The bands below are generated from the engine and apply to normalised FFMI, which adjusts for height so that tall and short people are compared fairly. Read them alongside a body fat figure: a fat percentage in the average band sitting next to an above-average FFMI describes a very different body from the same fat percentage sitting next to a below-average one.

• The male bands come from the sports-science and bodybuilding literature, where the upper figure of around 25 is frequently discussed as an approximate ceiling for drug-free trained men. That is a reported observation from small samples, not a law.
• The female bands are a working convention used to give women a comparable scale. They are not drawn from a published regression, and we would rather say so than imply a precision that does not exist.
• FFMI inherits the fat estimate's error. If the body fat figure is out by three points, the lean mass behind the FFMI is out too.

A chart is only as good as the figure you bring to it, and most figures cannot support the comparison being made with them. Three things decide whether yours can.

Method first. A DEXA scan is the practical reference in most settings, with air displacement and hydrostatic weighing comparable. Skinfolds in trained hands land within about three to five points and track change well. Tape-based formulas such as the Navy method carry a similar band and drift in the very lean and the very muscular. Consumer bioimpedance scales are the weakest of the group because they respond to hydration, and hydration moves by kilograms across a day.

Consistency second, and it matters more than method. Whatever you use, use it the same way every time: same time of day, same hydration state, same device, same operator, same side of the body for skinfolds, before rather than after training. A cheap method used consistently detects change far more reliably than an accurate method used haphazardly, because the systematic error largely cancels when you compare one of your own readings with another.

Time third. Body composition changes slowly, and the noise in any field method is larger than a month's honest progress. Comparing readings four to eight weeks apart is asking the method a question it cannot answer. Three to six months is a fair interval. Track the trend across several readings rather than reacting to the last one, and when the chart and the tape measure disagree, believe the tape measure; where fat sits tells you more about risk than how much of it there is.