How Accurate Is a TDEE Calculator?

A TDEE (Total Daily Energy Expenditure) calculator gives you a starting calorie target — the number of calories you burn in a day. The interesting question isn’t whether the calculator works. It’s how much you can trust the answer it gives you.

The short version: a TDEE calculator built on the Mifflin-St Jeor equation is accurate within roughly ±10% if you input honest numbers. That’s about ±200 kcal/day for the average adult — meaningful, but not enough to make the calculator useless. Most of the error doesn’t come from the formula. It comes from one place that has a fix.

Run our TDEE calculator first if you haven’t, then use this article to interpret the number it gave you.

What a TDEE calculator actually does

A TDEE calculator estimates your total daily energy expenditure by combining four things:

• Basal Metabolic Rate (BMR) — calories burned at rest just to keep your body running. Heart, brain, kidneys, cell repair. About 60-70% of TDEE for most adults. (BMR vs TDEE explained — the difference matters for setting calorie targets.)
• Thermic Effect of Food (TEF) — calories burned digesting and processing what you eat. About 8-12% of intake, varies by macro composition.
• Exercise Energy Expenditure (EEE) — workouts, runs, gym sessions.
• Non-Exercise Activity Thermogenesis (NEAT) — everything else: walking, fidgeting, typing, standing, talking with your hands.

Calculators don’t measure any of these directly. They estimate BMR with a formula based on age, sex, height, and weight, then multiply by an activity factor that’s supposed to capture EEE plus NEAT plus TEF in one number.

The most widely used BMR formula is Mifflin-St Jeor, published by Mifflin and St Jeor in 1990 in the American Journal of Clinical Nutrition (Mifflin et al., 1990). For men:

BMR = 10 × weight(kg) + 6.25 × height(cm) − 5 × age(y) + 5

For women:

BMR = 10 × weight(kg) + 6.25 × height(cm) − 5 × age(y) − 161

The activity multiplier ranges from about 1.2 (sedentary desk job) to 1.9 (athlete training twice a day). The framework dates back to Roza and Shizgal’s 1984 reevaluation of resting energy requirements (Roza & Shizgal, 1984).

Our TDEE calculator uses Mifflin-St Jeor with a standard 5-tier activity scale — the same setup recommended by the Academy of Nutrition and Dietetics.

The honest answer: ±10%, with caveats

A 2005 systematic review in the Journal of the American Dietetic Association compared the four most-used predictive equations against indirect calorimetry — the lab-grade gold standard for measuring resting metabolic rate. The conclusion:

The Mifflin-St Jeor equation was the most reliable, predicting RMR within 10% of measured in more nonobese and obese individuals than any other equation, and it also had the narrowest error range. (Frankenfield et al., 2005)

That’s the foundation of the ±10% number. If the formula says your TDEE is 2,200 kcal/day, your actual expenditure on a typical day is most likely between 1,980 and 2,420 kcal. The Academy of Nutrition and Dietetics formally recommends Mifflin-St Jeor on this basis.

A 2016 systematic review by Madden, Mulrooney, and Shah in the Journal of Human Nutrition and Dietetics extended the finding to obese populations specifically: Mifflin-St Jeor remained the most accurate, though the error band widens to roughly ±15% at BMI > 35 (Madden et al., 2016). For context on what calorie targets these numbers map to in practice, see how many calories per day — population-level needs from the USDA Dietary Guidelines.

What ±10% is not is a flaw of the formula. The 10% range reflects real biological variation between adults at the same age, sex, height, and weight. Two 35-year-old women, both 165 cm and 65 kg, can genuinely burn 1,650 vs 1,950 kcal/day at rest depending on lean mass, thyroid output, and other factors no calculator can see.

The good news: the part of the error you can fix is bigger than the part you can’t.

Where the error actually comes from

The total ±10% breaks into four sources, listed roughly by size.

BMR formula error (~3-7%)

The smallest piece. Mifflin-St Jeor uses age, sex, height, and weight as proxies for fat-free mass — the actual driver of resting metabolism. When body composition fits the population the formula was validated on (general healthy adults), error stays under 7%. For an athlete with much more lean mass than average, or an elderly adult with sarcopenia, the proxy breaks down and error climbs.

Activity multiplier error (~5-15%) — the biggest fixable piece

This is where most of the trouble is. The 5-tier activity scale (sedentary → light → moderate → very active → athlete) tries to compress a continuous reality into five buckets. Two adults in the same bucket can have wildly different real expenditure.

The dominant cause is NEAT. James Levine’s research at the Mayo Clinic showed NEAT can vary by approximately 2,000 kcal/day between adults of identical height, weight, and exercise habits — entirely from differences in how much they fidget, walk between rooms, talk with their hands, and stay on their feet (Levine, 2002). One person on the “moderately active” multiplier (1.55) genuinely burns 2,400 kcal; another at the same multiplier burns 2,000.

The activity multiplier also can’t see when you change. A new desk job, a longer commute, a winter month spent indoors — all change NEAT meaningfully without changing which bucket you tick.

Body composition mismatch (~2-5%)

If your body composition is far from average for your inputs, lean mass becomes the missing variable. The Katch-McArdle equation tries to fix this by using measured body fat percentage instead of weight alone — but only if the measurement is accurate. A bathroom scale’s body-fat estimate is not. A DEXA or Bod Pod scan is. For most people without recent DEXA data, sticking with Mifflin-St Jeor is more accurate than guessing body fat.

Thermic effect of food variability (~2-3%)

TEF averages ~10% of intake but ranges 5-15% by macro composition. Protein has the highest TEF (20-30%), carbs lower (5-10%), fat lowest (0-3%). A high-protein cutting diet pushes real expenditure 50-100 kcal/day above what the calculator predicts; a high-fat ketogenic diet pushes it slightly below. Calculators don’t see your macro split.

Mifflin-St Jeor vs Harris-Benedict — does it matter which one?

The Harris-Benedict equation dates to 1919 (Harris & Benedict). It was the standard for over 70 years before Mifflin-St Jeor replaced it. Many older calculators and textbooks still default to it.

The 2005 Frankenfield review put hard numbers on the comparison: Mifflin-St Jeor had a narrower error range than Harris-Benedict in both nonobese and obese populations. The gap is real but small — roughly 5% in the populations where it has been measured.

Why does Mifflin-St Jeor win? It was validated on a more representative modern population. Harris-Benedict was developed on younger, leaner adults from a century ago. Both formulas use the same inputs (age, sex, height, weight); Mifflin-St Jeor just has better-fit coefficients.

The takeaway: if your calculator uses Mifflin-St Jeor, you’re already on the most accurate option for the inputs you have. Switching to a different formula won’t close the ±10% gap meaningfully. What will close it is fixing the activity multiplier — and that’s a calibration problem, not a formula problem.

A note on Katch-McArdle: it can be marginally better than Mifflin-St Jeor if you have a recent, accurate body-fat measurement (DEXA, hydrostatic weighing, Bod Pod). With body-fat estimated by a bathroom scale or visual guess, Katch-McArdle is worse than Mifflin-St Jeor — you’ve added measurement error on top of formula error.

The 3-week self-calibration protocol

This is the part that closes the gap from ±10% to ±2%.

A calculator output is a hypothesis: “I think you burn 2,200 kcal/day.” You can test that hypothesis with three weeks of data and adjust until the number matches reality.

Step 1 — Day 1: Get the baseline

Run a calculator that uses Mifflin-St Jeor (any reputable one — ours is at /tools/tdee-calculator/). Be honest about activity level. If you’re not sure between two tiers, pick the lower one — most adults overestimate.

This is your starting target. Eat to it for the next three weeks. Don’t try to lose or gain weight yet — you’re calibrating, not dieting.

Step 2 — Days 1-21: Track honestly

Two pieces of data per day:

• Calorie intake. Track every meal. The closer to truth here, the more accurate the calibration. If precise tracking is unsustainable for you, see how to track calories without counting — photo-based methods are accurate enough for this purpose at a fraction of the effort.
• Body weight. Weigh yourself first thing in the morning, after using the bathroom, before eating or drinking. Same scale, same conditions, every day. Daily weight is noisy (water, glycogen, sodium, sleep), so don’t react to single-day numbers — what matters is the weekly average.

After 21 days you’ll have:

• Your average daily intake (kcal)
• Your week 1 average weight
• Your week 3 average weight

Step 3 — Day 22: Math

Weight change comes from energy balance. The thermodynamic rule of thumb: 1 kg of body fat ≈ 7,700 kcal (1 lb ≈ 3,500 kcal). So:

Energy gap (kcal/day) = (Week 3 weight − Week 1 weight) × 7,700 ÷ 21

Real TDEE = Average daily intake − Energy gap

A worked example. Calculator output: 2,200 kcal. You ate 2,200 every day for 21 days. Week 1 average weight: 70.5 kg. Week 3 average weight: 70.1 kg. Loss: 0.4 kg over 21 days.

Energy gap = 0.4 × 7,700 ÷ 21 ≈ 147 kcal/day

You lost weight at 147 kcal/day below maintenance. Your real TDEE was actually:

Real TDEE = 2,200 + 147 ≈ 2,347 kcal/day

The calculator was off by 147 kcal — well within the predicted ±10% range. Your new working number is 2,347.

Step 4 — Day 22 onward: Iterate

Run the same protocol with your new target. If a second three-week round shows weight stable within ±0.3 kg/week, you’ve converged on real TDEE within roughly ±50 kcal — about ±2%, the practical ceiling of self-experimentation without a metabolic chamber.

Re-run this once a quarter, or whenever your weight changes by 5+ kg, your activity habits shift, or your training load changes. Your TDEE is not a fixed number; it adjusts as your body and life change (Roza & Shizgal, 1984).

Once you have a calibrated TDEE, our calorie-deficit calculator maps it to a weight-loss target.

When TDEE calculators are especially wrong

There are populations where even calibrated calculators have wider error bands. Honesty matters more than optimism here.

• High-lean-mass athletes — formula proxies underestimate by 100-300 kcal. If you’re a serious lifter, runner, or competitive athlete, expect to calibrate to a higher number than the calculator gave you.
• BMI > 35 — every formula’s error widens. Mifflin-St Jeor remains the least bad, but expect ±15% rather than ±10% (Madden et al., 2016).
• Adults over 65 — formulas were validated mostly on younger populations. Metabolic rate at advanced age has more individual variance than any equation captures.
• Pregnancy and breastfeeding — calculator output is irrelevant. Energetics change in ways no general formula models. Work with a registered dietitian.
• History of disordered eating — a calorie target can re-trigger restrictive patterns. Do not self-calibrate; work with a clinician.
• Diagnosed metabolic conditions (thyroid disorders, PCOS, Cushing’s syndrome, type 2 diabetes on certain medications) — clinical estimation is needed. Formulas don’t model the disease effect.

If any of the above describes you, treat calculator output as discussion material with a healthcare provider, not as a working target. See our medical disclaimer for the full scope of what this article is and isn’t.

What about Apple Watch, Fitbit, and Whoop?

Wearables don’t replace calculators — they miss in different ways.

Heart-rate-based exercise estimates over-predict workout calorie burn substantially. A 2017 Stanford study tested seven popular wrist-worn devices (Apple Watch, Fitbit, Samsung Gear, etc.) against indirect calorimetry across 60 participants and found that none had energy-expenditure error under 27%, with a median error of about 27% across the cohort and individual-device errors as high as 92.6% (Shcherbina et al., 2017). The error is largest for steady-state cardio and smallest for short, hard intervals.

Step-counter algorithms only see locomotion. They miss most of NEAT — typing, talking with your hands, fidgeting, standing. The gap between what your watch credits you with and what you actually burn at rest can be ±300 kcal/day.

The pragmatic take: a wearable is one signal, a calibrated calculator is another, and the scale-based 3-week calibration is a third. If all three roughly agree, your number is solid. If they disagree, trust the scale — it’s the only one of the three that measures the actual outcome (weight change), not an estimate of inputs.

What actually beats a calculator

Three measurement options exist for adults willing to spend money. Each has a sharp limitation.

• Indirect calorimetry — the gold standard for measuring BMR specifically. You sit still under a metabolic hood for 20-30 minutes; the device measures oxygen consumed and CO₂ produced and calculates resting metabolism. Available at most sports-medicine and weight-management clinics in the US. Cost: $80-200 per session. Limit: only measures BMR, not full TDEE — you still need an activity multiplier for the rest.
• Doubly labelled water (DLW) — the gold standard for measuring TDEE itself. You drink isotopically labeled water, urine samples are collected over 7-14 days, and the elimination rates reveal your real TDEE under free-living conditions (Westerterp, 2017). Limit: research-only, costs roughly $500-1,000 per test, requires a research lab. Almost no consumer access.
• DEXA scan — measures body composition (lean mass, fat mass, bone). Refines the inputs to Katch-McArdle. Cost: $50-150 per scan. Limit: gives you better inputs, not a TDEE measurement.

For 99% of adults, a scale plus 21 days of honest tracking outperforms any of these for the cost. Indirect calorimetry is useful if you have reason to suspect your BMR is unusual — recovery from very low intake, suspected metabolic adaptation, or a specific medical question. For most people, it’s spending $80-200 to learn what three weeks of weighing yourself would tell you for free.

In one line

A TDEE calculator gets you to ±10% in 30 seconds. Three weeks of weighing yourself gets you to ±2%. Both numbers are useful. Only one of them is yours.

Start with our TDEE calculator, or skip the manual logging entirely — send a meal photo to Nouri in Telegram and the food side of the calibration takes care of itself.

Sources

1. Frankenfield D, Roth-Yousey L, Compher C. Comparison of predictive equations for resting metabolic rate in healthy nonobese and obese adults: a systematic review. J Am Diet Assoc 2005;105(5):775-89. PubMed 15883556
2. Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr 1990;51(2):241-7.
3. Harris JA, Benedict FG. A Biometric Study of Basal Metabolism in Man. Carnegie Institution of Washington, 1919.
4. Roza AM, Shizgal HM. The Harris Benedict equation reevaluated: resting energy requirements and the body cell mass. Am J Clin Nutr 1984;40(1):168-82.
5. Madden AM, Mulrooney HM, Shah S. Estimation of energy expenditure using prediction equations in overweight and obese adults: a systematic review. J Hum Nutr Diet 2016;29(4):458-76.
6. Levine JA. Non-exercise activity thermogenesis (NEAT). Best Pract Res Clin Endocrinol Metab 2002;16(4):679-702.
7. Westerterp KR. Doubly labelled water assessment of energy expenditure: principle, practice, and promise. Eur J Appl Physiol 2017;117(7):1277-1285.
8. Shcherbina A, Mattsson CM, Waggott D, et al. Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort. J Pers Med 2017;7(2):3. PubMed 28538708
9. Academy of Nutrition and Dietetics — Evidence Analysis Library. Resting Metabolic Rate (RMR) Equations.

Frequently asked

How accurate is the Mifflin-St Jeor equation?

Mifflin-St Jeor predicts resting metabolic rate within ±10% in roughly 70-80% of healthy adults — the narrowest error range of any widely-used predictive equation, per Frankenfield et al., 2005 (J Am Diet Assoc). Accuracy widens to ±15% at BMI > 35 (Madden et al., 2016).

Mifflin-St Jeor vs Harris-Benedict — which is more accurate?

Mifflin-St Jeor is approximately 5% more accurate on average, both for nonobese and obese adults. Harris-Benedict was validated on a younger, leaner 1919 population and tends to overestimate slightly. Switch any calculator that defaults to Harris-Benedict to Mifflin-St Jeor if you can.

Why is my TDEE so low?

Three common reasons. First, activity level was set higher than reality — most adults overestimate. Try the next-lower tier. Second, body composition skews lean — you have less muscle than the formula assumes for your weight, common after long calorie restriction. Third, an input mix-up — height in cm vs inches, weight in kg vs lbs.

Can a TDEE calculator be off by 500 calories?

Yes, especially when the activity multiplier doesn’t match real non-exercise activity. James Levine’s research showed NEAT alone varies by approximately 2,000 kcal/day between adults of identical stats. Two people who tick the same activity tier can genuinely differ by 300-500 kcal/day in real expenditure (Levine, 2002).

How often should I recalculate my TDEE?

Every 5-10 kg of weight change, every 4-6 weeks during active dieting or bulking, or whenever your activity habits shift meaningfully (new job, new training program, season change). TDEE moves with body mass — a 10 kg loss drops TDEE by roughly 100-150 kcal/day.

Is BMR more accurate than TDEE?

BMR is measured more accurately because indirect calorimetry is widely available; TDEE is estimated from BMR plus an activity multiplier, and the multiplier is where most error enters. Measured BMR has tighter error than calculated TDEE, but for daily decisions about how much to eat, TDEE is the relevant number.

Why does MyFitnessPal give a different TDEE than other calculators?

Different apps use different default activity multipliers and sometimes different BMR formulas. MyFitnessPal also factors in exercise logged into the app — adding workouts on top of an already-set baseline. If two apps disagree by less than ~15%, both are within normal calculator variance; pick one and stick with it for consistency.

How can I make my TDEE estimate more accurate?

Run the 3-week self-calibration protocol described in this article: eat at the calculator number, track intake honestly, weigh daily, average weekly, then back-calculate real TDEE from weight change × 7,700 kcal/kg. This brings accuracy from ±10% to about ±2% for almost anyone without a medical condition.