What Is BMI and How Accurate Is It?

BMI is accurate as a population-level screening tool, but it's far less accurate for individuals. The Body Mass Index—that simple formula of weight in kilograms divided by height in metres squared—ranks among medicine's most useful and most misunderstood metrics. This guide explains what BMI actually measures, where it falls short, and what other numbers matter more for your health.

What BMI Is (and Isn't)

The formula is straightforward: BMI = weight (kg) ÷ height (m)². The WHO classifies adults as:

Underweight: BMI under 18.5
Healthy weight: BMI 18.5–24.9
Overweight: BMI 25–29.9
Obese: BMI 30 or above

The NHS uses these same bands, and they're based on epidemiological data linking BMI to mortality risk across millions of people. At the population level, this works: countries with higher average BMI generally have higher rates of cardiovascular disease, type 2 diabetes, and certain cancers.

But here's the critical bit: BMI is a screening tool, not a diagnosis. A BMI of 26 doesn't mean you're unhealthy. A BMI of 22 doesn't guarantee you are. The number is a flag to consider—not a verdict.

Calculate yours with our BMI calculator for a baseline reading.

How Accurate Is BMI, Really?

The accuracy question has a straightforward answer: it depends entirely on who you are.

BMI is reasonably accurate for sedentary populations with typical body composition. It's problematic for several groups:

Muscular people and athletes. Muscle weighs more than fat, so a 90kg powerlifter with 12% body fat will have a BMI of 29–31 (technically "overweight" or "obese") despite excellent health markers. Muscular individuals and athletes face particular challenges when interpreting BMI.
Elderly populations. Bone density decreases with age. A 80-year-old with a BMI of 26 has less muscle and more fat than a 35-year-old with the same BMI, yet both fall in the "healthy" range.
Children. Growth patterns are non-linear. BMI works differently for children, using age-adjusted percentile charts instead of fixed bands. What's healthy at age 8 is very different at age 14.
Extreme BMI values. Below 15 or above 40, BMI becomes increasingly inaccurate as a health predictor.

The deeper issue: BMI conflates two entirely different things—total body weight and height. It doesn't distinguish between fat and muscle, bone density, organ size, or water retention. Two people with identical BMI can have radically different body compositions. This matters. Some research suggests BMI misclassifies a meaningful proportion of adults when compared against actual body composition, particularly among athletic populations.

What Matters More Than BMI

If BMI is this imprecise, what should you actually measure?

Body fat percentage is far more predictive of health outcomes. Research consistently shows that body composition—the ratio of muscle to fat—predicts cardiovascular risk better than BMI does. Here's what the numbers mean for your health. Healthy ranges:

Men: 10–20%
Women: 18–28%

You can estimate this with online calculators, DEXA scans (most accurate, ~£100), or skinfold calipers (free but technique-dependent).

Waist circumference is surprisingly predictive. NICE obesity guidance notes that waist circumference independently predicts metabolic syndrome and cardiovascular risk. Measure at your navel (standing relaxed, not sucked in):

Men: keep under 94cm
Women: keep under 80cm

You can have a BMI of 23 and elevated cardiometabolic risk if your waist circumference is 100cm. Conversely, someone with a BMI of 28 and a waist of 85cm may have lower metabolic risk. This single measurement often matters more than the BMI number itself.

Muscle-to-fat ratio is essentially body composition viewed through a different lens. Understanding muscle-to-fat ratio helps you see why two people of identical weight, height, and BMI can look and perform completely differently. One might be lean and muscular; the other might be soft and sedentary.

When BMI Still Works

Despite its limitations, BMI remains useful for:

Population health trends. If a country's average BMI climbs from 25 to 27 over a decade, something has shifted—usually reduced activity combined with increased caloric intake.
People with sedentary lifestyles. If you don't lift weights and don't have unusual muscle mass, BMI is a reasonable estimate of whether your weight is reasonable for your height.
Longitudinal personal tracking. If your BMI has climbed from 24 to 27 to 30 over three years, that's a real signal of change, even if the absolute number alone isn't a diagnosis.

The NHS and government guidance still use BMI because it's free, quick, repeatable, and works well enough for population-level decisions. It's not perfect—but it's better than guesswork.

Context Changes Everything

How to interpret your BMI depends on several factors:

Age. As you age, you naturally lose muscle mass and bone density (unless you strength train). How your age affects calorie needs and metabolism explains why a BMI of 26 might be perfectly healthy at 65 but concerning at 25.

Genetics. Some people naturally carry more muscle (genetic predisposition to lean mass), while others naturally carry more fat. A BMI of 27 might be your body's set point; for someone else, it's genuinely concerning.

Training status. Strength athletes, rowers, and rugby players routinely have BMI values of 28–32 with single-digit body fat. Sedentary people can have a "healthy" BMI of 22 with 28% body fat—technically overfat despite looking unremarkable clothed.

Medications and health conditions. Steroids, hormonal medications, and conditions like PCOS or hypothyroidism affect weight independently of calories. BMI can't capture these contexts.

Using BMI Sensibly

The most practical approach uses BMI as one data point among several:

Get a baseline BMI using our BMI calculator.
Measure your waist at your navel while standing—this single number often predicts health better than BMI.
Consider body composition. If you lift weights, estimate your body fat percentage or get a DEXA scan.
Track trends, not absolutes. Recheck every 4–8 weeks and watch the slope of your trend line, not individual measurements.
Link to broader health. BMI is one piece of the puzzle alongside activity level, sleep, stress, cardiovascular fitness, and bloodwork.

If you're making dietary or exercise changes, understanding your ideal weight alongside BMI gives you a fuller picture of what healthy looks like for you.

Frequently Asked Questions

Is a BMI of 25 really overweight? Technically yes by WHO definition, but context matters entirely. If you're 25 with 15% body fat, you're lean and muscular. If you're 25 with 32% body fat, you're overfat despite the "normal" number. The BMI band is a starting point for conversation, not a verdict.

Does my BMI matter if I'm fit? Not particularly. Cardiovascular fitness (VO2 max, resting heart rate) is a far better predictor of longevity and disease risk than BMI. Elite athletes regularly have "overweight" or "obese" BMI values. If you can run a 10km without stopping or deadlift 1.5× your body weight, your BMI number is largely irrelevant.

Why don't doctors use body fat percentage instead? Because body fat requires special equipment—DEXA scans cost money, calipers require training. BMI is free and takes 10 seconds. For a GP doing health checks on 50 people a week, BMI is pragmatic even if it's not perfect. It's a first filter: if your BMI is 22, they don't investigate further; if it's 32, they ask questions.

Can I have a "healthy" BMI and still be unhealthy? Yes. You can be thin-fat—a BMI of 22 with 32% body fat, thanks to sedentary lifestyle. Conversely, you can be overfat by BMI but very fit thanks to training. This is why waist circumference matters; it catches the thin-fat problem that BMI misses.

What's the ideal BMI? For most people, 20–23 is a reasonable target if you're also strength training (so you're carrying muscle). But "ideal" depends on your genetics, training status, and personal health markers. A BMI of 24 with visible muscle definition and good bloodwork is far better than a BMI of 21 from crash dieting and no exercise.

Should I use BMI for my children? No—not the adult bands. BMI for children uses age-adjusted percentile charts instead of fixed thresholds. A BMI that's "overweight" for a 7-year-old might be normal for a 14-year-old.

Does weight training mess up my BMI? Yes, by design. If you're building muscle, your BMI might stay the same or even climb while your body fat drops. This is why BMI is a terrible metric for tracking progress in a strength program—use photos and body measurements instead.

Is there a BMI that's too low? Yes. BMI below 18.5 is classified as underweight, and prolonged underweight status (BMI under 17) is associated with bone density loss, immune suppression, and other health risks. But like the upper end, context matters. A very lean athlete with 8% body fat and excellent bone density is different from someone underfed through restriction.