How accurate is a TDEE calculator?

Close enough to start with, wrong often enough that you should plan to replace it with your own data.

Written and checked byMadhu Kinnera · Last reviewed September 1, 2026

In short

For most healthy adults a good BMR equation lands within about 10% of a measured value, roughly ±150 kcal at the resting step alone.
Around one person in five is missed by more than that, and a calculator cannot tell you which group you are in.
Which activity multiplier you pick changes the answer more than which formula the calculator uses.
The single largest error source is non-exercise movement, which varies by hundreds of calories a day between similar people.
The fix is not a better calculator. It is two to three weeks of your own weight data, then a small adjustment.

How accurate is a TDEE calculator, really?

A TDEE calculator is accurate enough to be a useful starting point and not accurate enough to be a target you chase to the calorie. The best-validated equation for the general population, Mifflin-St Jeor, predicts resting metabolic rate to within 10% of a measured value for about 80% of healthy non-obese adults and about 70% of people with obesity. That is the good news. The other side of it is that roughly one person in five is missed by more than 10%, sometimes far more, and no calculator can tell you which group you are in from six input fields.

Ten percent does not sound like much until you attach a number to it. A resting rate of 1,600 kcal carries a ±160 kcal band before anything else is added, and that is only the first step of the calculation. The activity multiplier applied on top of it introduces its own, larger error. So "is a TDEE calculator accurate?" has a precise answer: it is a good estimate with roughly ±16% of real uncertainty around it, and any tool that hands you a single unqualified number is hiding that from you.

The formula is almost never the problem

People searching for the most accurate TDEE calculator usually assume the differences between tools come down to which equation each one uses. They rarely do. The Mifflin-St Jeor calculator and the revised Harris-Benedict equation typically agree to within 50–80 kcal for the same person, and Frankenfield's systematic review found Mifflin-St Jeor to be the most reliable of the common equations, with the narrowest error range. Switching between reputable formulas moves your estimate by less than a snack.

The Katch-McArdle equation is the one real exception, and only in one direction. Because it works from lean body mass instead of total weight, it is more accurate than Mifflin-St Jeor for lean, muscular, or unusually heavy people, but only when the body-fat percentage feeding it is a real measurement. With a guessed body-fat number it is less accurate, while looking more precise. This is covered in full on the Katch-McArdle calculator page.

Why two TDEE calculators give you different numbers

Run the same details through five sites and you can get five answers 300–400 kcal apart. Almost all of that spread comes from two places, and neither is the formula:

The activity multiplier. Moving one bucket (say "lightly active" to "moderately active") multiplies your resting rate by a factor 0.15–0.20 higher, which is 150–250 kcal for most people. Different sites also label these buckets differently, so the same description of your week maps to different multipliers.
Rounding and assumptions. Some tools round the multiplier, some bake in a small activity allowance, some estimate body fat when you leave it blank. Each choice is defensible and each one shifts the output.

The difference between BMR and TDEE is exactly this multiplier step, and it is where the uncertainty lives. A calculator that shows you the multiplier it used, and lets you see what a different one would produce, is more honest than one that picks silently.

The biggest single source of error is movement you never notice

Non-exercise activity thermogenesis (NEAT) is every calorie you burn that is not sleeping, digesting, or deliberate exercise: walking to the kitchen, standing, fidgeting, taking the stairs, gesturing while you talk. It is also the most variable component of energy expenditure between people.

In Levine's overfeeding study, participants were fed 1,000 kcal a day above maintenance for eight weeks. The change in their NEAT ranged from a small decrease to an increase of nearly 700 kcal a day, and that difference alone predicted a tenfold spread in how much fat each person stored. Two people with the same weight, age, sex and stated gym schedule can genuinely differ by several hundred calories a day in what they burn just living, and no equation built from height and weight can see any of it.

Why two identical-looking people burn different amounts

Even at complete rest, people matched on height, weight, age and sex differ in metabolic rate: the coefficient of variation in resting metabolic rate runs roughly 5–10%, and that residual variation persists after body composition is accounted for. Some of it is organ size, some is hormonal, some is not well explained at all. A prediction equation returns the population average for your inputs; your own rate is that average plus or minus an amount the equation cannot know.

This is why the calculator here draws a band rather than a point. The band is not decoration and it is not hedging: it is the actual width of what the evidence supports, and your true figure is somewhere inside it.

How to find your real TDEE

You measure it, and it takes about three weeks:

Take the estimate from any reasonable calculator, this one included.
Eat as close to that number as you can manage, every day, for two to three weeks. Weigh and log honestly: under-reporting intake is common and it will corrupt the result.
Weigh yourself each morning under the same conditions and track the weekly average, not individual days. Day-to-day swings are water and gut contents and will mislead you.
If your weekly average held steady, the estimate was close. If it rose, your real maintenance is lower than the estimate: subtract about 100–150 kcal. If it fell, add the same. Hold the new number for another two weeks and check again.

Two rounds of that will beat every equation on the internet, because your own weight trend reflects your NEAT, your adaptation and your real adherence: the things no formula can observe. At that point you have measured your TDEE instead of predicting it, and you no longer need a calculator for it.

What this calculator does about all this

It shows the range instead of hiding it. The TDEE calculator on the homepage draws a ±16% band around your estimate, narrowing to about ±11% if you enter a measured body-fat percentage, and marks your point estimate inside it. It names the formula it used, substitutes your own numbers into the arithmetic, and shows the exact activity multiplier it applied so you can judge whether it matches your week.

None of that makes the estimate more accurate than the evidence allows. It makes the estimate honest, which is the part that lets you use it well: you know what you are starting from, you know how rough it is, and you know the next step is to replace it with your own data.

Sources

The accuracy figures on this page come from the primary literature, not from other calculators:

Frankenfield D, Roth-Yousey L, Compher C (2005). Comparison of predictive equations for resting metabolic rate in healthy nonobese and obese adults: a systematic review. Journal of the American Dietetic Association, 105(5), 775–789.

Mifflin-St Jeor as the most reliable of the commonly used equations, with the narrowest error range.
Frankenfield D, Rowe WA, Smith JS, Cooney RN (2003). Validation of several established equations for resting metabolic rate in obese and nonobese people. Journal of the American Dietetic Association, 103(9), 1152–1159.

The 80% / 70% within-10% figures for nonobese and obese adults.
Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO (1990). A new predictive equation for resting energy expenditure in healthy individuals. American Journal of Clinical Nutrition, 51(2), 241–247.

The default BMR equation and the population it was validated on.
Levine JA, Eberhardt NL, Jensen MD (1999). Role of nonexercise activity thermogenesis in resistance to fat gain in humans. Science, 283(5399), 212–214.

The between-person spread in NEAT under overfeeding and its link to fat gain.
Donahoo WT, Levine JA, Melanson EL (2004). Variability in energy expenditure and its components. Current Opinion in Clinical Nutrition and Metabolic Care, 7(6), 599–605.

The 5–10% coefficient of variation in resting metabolic rate between matched individuals.

Medical disclaimer

This page is written by Madhu Kinnera, who is not a clinician. Nothing here is medical advice. If you have a condition that affects your metabolism, are pregnant, are recovering from an eating disorder, or are planning a large change in how much you eat, talk to a doctor or a registered dietitian before acting on any calculator, this one included.

Frequently asked questions

How accurate is a TDEE calculator?

For most healthy adults, a good BMR equation lands within about 10% of a measured resting rate, and the activity multiplier adds error on top of that, so treat the output as carrying roughly ±16% of real uncertainty. About one person in five is missed by more than 10% at the resting step alone, and the calculator cannot tell you whether you are one of them.

Which is the most accurate TDEE calculator?

No calculator is meaningfully more accurate than another if they use a well-validated equation, because the formula is not where the error is. What separates a good tool from a bad one is honesty: whether it shows the range, names the multiplier it used, and tells you how to replace the estimate with your own data. A confident single number is a worse tool, not a more accurate one.

Why do different TDEE calculators give me different results?

Almost entirely because they assume different activity multipliers or round them differently. Moving one activity bucket changes the result by 150–250 kcal for most people, which dwarfs the 50–80 kcal difference between reputable BMR formulas. Rounding, hidden activity allowances, and how each tool handles a blank body-fat field account for the rest.

Is the Mifflin-St Jeor equation accurate?

It is the most reliable of the commonly used equations for the general population: the narrowest error range of the four Frankenfield’s 2005 review compared, and validation work puts it within 10% of measured resting rate for about 80% of non-obese adults. It is still a population average, so it will be off for you by some unknown amount, which is why this site shows a band around it rather than a single figure.

How can I find my real TDEE?

Eat a fixed, honestly logged intake for two to three weeks, weigh yourself every morning, and compare weekly averages rather than individual days. If your average weight held steady, the intake you ate is your maintenance. If it moved, adjust by about 100–150 kcal and repeat. Your own weight trend is a measurement; a calculator is only a prediction.

Why is my TDEE calculator so far off?

Usually one of three things: you chose an activity level a bucket too high, you are in the minority whose resting rate the equation misses by more than 10%, or your non-exercise movement is well above or below what the multiplier assumes. Under-reported food intake is the fourth common cause when the number seems too high. The remedy for all four is the same three-week weigh-and-adjust process.

Does body fat percentage make a TDEE calculator more accurate?

Only if it is measured: by DEXA, a skinfold assessment, or a clinical scan. A real figure lets the tool use the Katch-McArdle equation, which reads leanness directly and narrows the band from about ±16% to ±11%. A guessed body-fat number makes the estimate worse while making it look more precise, so leave the field blank unless you have a real measurement.

Are online calorie calculators accurate for weight loss?

They are accurate enough to set a sensible starting deficit (10–25% below the estimate) but not accurate enough to explain a stall to the calorie. If weight loss slows, the fix is to compare three weeks of weekly weight averages against your intake and adjust from what actually happened, not to recalculate and trust a new number.

How much can a TDEE estimate realistically be wrong by?

At the resting step, roughly ±10% for most people and more for about one in five. The activity multiplier can add another 150–250 kcal of error in either direction. Combined, an individual estimate being 300–500 kcal off maintenance is well within normal, which is why a single unqualified number is misleading.

What is NEAT and why does it matter for accuracy?

NEAT (non-exercise activity thermogenesis) is everything you burn through incidental movement: walking, standing, fidgeting, posture. It varies by several hundred calories a day between people with the same weight and stated activity level, and it adapts when you eat more or less. It is the single largest reason two similar people have different real TDEEs, and no height-and-weight equation can estimate it.

Should I use the higher or lower number if two calculators disagree?

Use the lower one as your starting intake. Correcting upward from a slight underestimate (you lose a little faster than planned, so you add calories) is straightforward. Correcting downward from an overestimate means diagnosing why a deficit that should work is not working, which is harder and more discouraging.

Does TDEE accuracy get worse as I lose weight?

The estimate needs updating more often. A smaller body burns less, and incidental movement often drops during a diet too, so a figure that was accurate at the start drifts over a long cut. Recalculate after every 5–10 kg of change, and lean on your weekly weight trend between recalculations.

Is a TDEE calculator accurate for athletes or very muscular people?

Less so with a weight-based equation, because it cannot see that more of your weight is metabolically active tissue. If you have a measured body-fat percentage, the Katch-McArdle equation handles this better. Even then, high training volumes make exercise energy expenditure hard to pin down, so the weigh-and-adjust process matters more, not less.