How Accurate Is a TDEE Estimate?

The short answer

A TDEE estimate is usually within about 10% of measured energy expenditure for a healthy adult, and the activity multiplier adds a further 5% to 15% of uncertainty. On a 2,600 kcal maintenance figure, that means the true number could plausibly sit anywhere between about 2,200 and 3,000 kcal.

That sounds discouraging, and it is the most useful thing on this page. A 400 kcal band is wide enough to explain why someone following the number exactly still does not lose weight, and it is why the last section of every sensible calorie conversation is the same: adjust from your own results rather than defending the equation.

Where the error comes from

The equations were derived by measuring resting energy in a few hundred to a few thousand people and fitting a line to age, sex, height and weight. That line explains most of the variation between people, but individual physiology - thyroid function, body composition, muscle efficiency, gut microbiome - sits in the residual.

Body composition is the biggest single factor the equations ignore. Two people of the same weight and height can differ by several kilograms of lean mass, and lean tissue costs energy at rest while fat tissue barely does. That is why a trained lifter often measures higher than the equation predicts and someone with low muscle mass measures lower.

The activity multiplier is the weakest link

Most people overestimate their activity level. Training three times a week while sitting for nine hours a day is not 'very active', and the difference between the 1.375 and 1.725 multipliers is more than 500 kcal a day at a 2,600 kcal baseline. Choosing the higher figure and eating to it is one of the most common reasons a diet stalls.

A more reliable approach is to pick the level that matches your last month rather than your best week, and to use step counts as a tie-breaker: under 5,000 steps is sedentary, 5,000 to 7,500 is light, 7,500 to 10,000 is moderate, and above 10,000 with regular training is very active.

Wearables do not solve the problem

Consumer devices estimate energy from heart rate, motion and sometimes temperature, and validation studies consistently find errors of 10% to 30% for daily totals. The pattern of error is predictable: reasonable accuracy for steady-state running, poor accuracy for strength training, intervals and short bouts, because heart rate lags behind the work and stays elevated afterwards.

Devices are excellent for relative tracking - comparing today with last month under similar conditions - and mediocre for absolute numbers. Using one to set a food target imports its error directly into your plan.

Turning a 10% error into a working plan

The fix is not a better equation; it is a feedback loop. Start with the estimate, hold intake steady for two weeks, weigh yourself daily under identical conditions, and compare weekly averages. A flat average means the estimate was right. A falling average means your true maintenance is higher than estimated - add 100 to 200 kcal. A rising average means the opposite.

Two or three rounds of this converge on a personal number that includes everything the equations miss: your job, your commute, your steps, your training and your digestion. That number is worth more than any formula, and it is the only one that will still be accurate in six months.

• Hold intake constant for a full two weeks before judging anything.
• Weigh daily, but compare seven-day averages only.
• Change one thing at a time, then wait two weeks again.
• Re-estimate after every 5 to 7 kg of weight change.

How much precision do you actually need

Less than the internet suggests. A 10% error in a calorie target is smaller than the difference made by two weeks of consistent adherence, and far smaller than the difference between tracking and not tracking. Chasing a 50 kcal improvement in the equation while missing the weekend days in your food log is optimising the wrong thing.

Where precision does matter is the direction and size of the trend. Week-to-week accuracy of your own weigh-in trend is high, which is why the feedback loop beats the formula every time, and why anyone who has been tracking for three months knows their true maintenance better than any calculator can.

Four cheap ways to sharpen the estimate

Measure at the same time of day, after the bathroom and before eating. Morning weight is the least noisy signal you can collect for free.

Log two normal weeks rather than two good ones. A week of perfect logging understates the weekends that follow it, and the average is what sets your maintenance.

• Use a 7-day average of weight, never a single morning
• Hold activity steady while you test, so the multiplier stops moving
• Count drinks and cooking oil, the two most under-reported items
• Re-run the numbers after every 5 kg of weight change

Frequently asked questions