How Accurate Is the TDEE Formula?
The Mifflin-St Jeor equation is the most validated formula for predicting resting metabolic rate (RMR). A 2005 meta-analysis by Frankenfield et al. tested five predictive equations against measured RMR in 2,000+ subjects and found Mifflin-St Jeor was accurate within ±10% for roughly 82% of the non-obese population - the best performance of any widely used formula.
In practice, that means:
Actual TDEE range at ±10%: 2,340–2,860 kcal
That's a 520 kcal spread. Significant - but still a useful starting point that gets you in the right zone.
The formula performs worse for specific groups:
- Very high or very low body fat: The equation uses total body weight, not lean mass. A 90 kg person at 8% body fat has a very different metabolism than a 90 kg person at 35% body fat. Katch-McArdle (which uses lean body mass) is more accurate here.
- Older adults: Metabolic rate declines with age beyond what weight alone predicts. Mifflin-St Jeor adjusts for age, but still tends to overestimate RMR in adults over 70.
- Extreme training volume: The activity multipliers were built on normal populations, not competitive athletes training 15+ hours per week.
For everyone else - the majority of gym-going adults - the formula is solid. Use the TDEE calculator to get your starting number, then validate it with 2 weeks of real weight data.
The Activity Multiplier Problem
This is where most formula errors originate. The activity multiplier converts your RMR into TDEE. Pick the wrong category and your entire calorie target shifts by 200–400 kcal/day before you've eaten a single meal.
The standard descriptions are vague. Here's what each level actually looks like in practice:
| Level | Standard Description | What It Actually Means | Multiplier |
|---|---|---|---|
| Sedentary | Little or no exercise | Desk job, walks fewer than 5,000 steps/day, no structured training | 1.2 |
| Lightly Active | 1–3 days exercise/week | 3 gym sessions/week, desk job, 5,000–8,000 steps/day | 1.375 |
| Moderately Active | 3–5 days exercise/week | 4–5 gym sessions/week + active daily movement, 8,000–12,000 steps | 1.55 |
| Very Active | 6–7 days exercise/week | Daily training + active job (teacher, nurse), 12,000+ steps | 1.725 |
| Extra Active | Physical job + training | Construction worker, athlete, or military with daily hard training | 1.9 |
A desk job with 3 gym sessions per week is Lightly Active (1.375) - not Moderately Active (1.55). That single category error creates a 290 kcal/day phantom surplus for an 80 kg male. Over 30 days: 8,700 kcal - roughly 1.1 kg (2.4 lb) of fat not lost, on paper, before you've even tracked food.
When in doubt, select one category lower than you think. Adjust based on what the scale does over two weeks - not based on how hard your workouts feel.
Why Your Tracking Is Off (More Than the Formula)
The formula gets blamed. The tracking is usually the problem.
Two landmark studies quantify this. In 1992, Lichtman et al. (published in The New England Journal of Medicine) studied 10 obese participants who reported "eating very little" but couldn't lose weight. Under metabolic ward conditions with doubly labeled water to measure actual intake:
- Self-reported intake: 1,028 kcal/day
- Actual measured intake: 2,081 kcal/day
- Average underreport: 47%
A 2015 analysis by Dhurandhar et al. found that even motivated, health-focused participants underreported by an average of 429 kcal/day when using standard food logging methods.
Where do the hidden calories come from? Four sources account for most of it:
Cooking Oil & Butter
1 tablespoon of olive oil = 120 kcal. Two tablespoons per day - a light hand by most cooking standards - adds 240 kcal that rarely makes it into the log.
Portion Size Estimation
A "medium banana" in MyFitnessPal is logged as 105 kcal. An actual medium banana weighs 120–130g and runs 120–140 kcal. Multiply small discrepancies across every meal and the gap grows fast.
Tasting While Cooking
Bites and tastes while preparing food average 100–200 kcal/day for people who cook regularly. Nothing gets logged. Nothing gets counted.
Drinks
Two lattes with whole milk = ~300 kcal. Fruit juice, protein shakes measured by eye, alcohol - all easy to undercount. "Just coffee" is rarely just coffee.
The formula gives you ±10% error. Tracking gives you ±20–30% error. Fix your tracking before adjusting the formula. The formula is almost certainly not the problem.
Real Comparison - Formula vs Actual (3 Scenarios)
The following scenarios use the Mifflin-St Jeor formula applied to real-world activity profiles, then compared to the maintenance intake that would be inferred from stable weight over an 8-week tracking period.
| Scenario | Profile | Formula TDEE | Actual Maintenance | Accuracy |
|---|---|---|---|---|
| A - Desk job, 3×/week gym | 80 kg (176 lb), Lightly Active (1.375) | 2,310 kcal | 2,250–2,350 kcal | ✅ Within 3% |
| A (miscategorized) | Same person, selected Moderate (1.55) | 2,600 kcal | 2,250–2,350 kcal | ❌ Off by ~280 kcal |
| B - Teacher, stands all day, 4×/week gym | 75 kg (165 lb) female, Very Active (1.725) | 2,890 kcal | 2,600–2,700 kcal | ⚠️ Off by ~200 kcal (7%) |
| C - Competitive athlete, 6×/week training | 85 kg (187 lb) male, Extra Active (1.9) | 3,400 kcal | 3,200–3,500 kcal | ✅ Within range |
Scenario B is instructive. "Standing all day" sounds like high activity - but standing burns only marginally more than sitting. The "Very Active" multiplier assumes continuous physical movement, not standing. A teacher on their feet is more accurately Moderately Active (1.55), bringing the TDEE down to ~2,600 kcal and aligning with actual maintenance.
The takeaway: the formula works. Activity level selection is where most people go wrong.
How to Know If Your TDEE Is Wrong
Stop guessing. Run the test.
Eat at your calculated TDEE for two weeks - no intentional deficit or surplus. Weigh yourself every morning at the same time, before eating or drinking. After 14 days, compare the weekly averages:
Weight Stable (±0.3 kg)
Formula is accurate for you. You can now adjust from this baseline with confidence.
Weight Dropping
Actual TDEE is higher than calculated. Add 100–150 kcal and retest. Don't jump straight to large increases.
Weight Rising
Actual TDEE is lower - or tracking is off. Before reducing calories, spend one week logging everything meticulously including oils, drinks, and bites. Fix tracking first.
Adjust by 100–150 kcal at a time. Not 300 or 500. Small adjustments based on 2-week averages give you data. Large adjustments based on a bad week give you noise.
Recalculate your TDEE whenever you adjust activity level, and use the calorie calculator to set exact fat loss or bulking targets from your validated maintenance number.
When to Recalculate
Your TDEE is not a fixed number. It changes as you change.
- Every 4–6 weeks during a cut or bulk: Every 5 kg (11 lb) of weight lost shifts your TDEE down by 50–100 kcal. Ignore this and progress stalls after the first month.
- After a 5 kg weight change in either direction: Both weight loss and significant muscle gain shift your maintenance calorie needs.
- After a major activity change: New job, injury, new training program, or any shift in daily movement patterns.
- After a long diet break or reverse diet: Extended caloric restriction causes metabolic adaptation. Your maintenance intake post-diet may be lower than it was pre-diet even at the same weight.
Your TDEE from 6 months ago is not your TDEE today.
For setting calorie targets after recalculating, see how to structure a lean bulk at the right surplus, or use the specific calorie targets in the lean bulk calories guide for bodyweight-based examples.