Nutrition

How to Calculate Your Maintenance Calories Accurately

The Mifflin-St Jeor formula is a starting estimate — it’s wrong by 15–20% for most people. The formula gives you a starting point. Your actual maintenance is what your scale tells you over 3 weeks.

The short answer

To estimate your maintenance calories, multiply your Mifflin-St Jeor BMR by your activity multiplier (1.2 to 1.9 depending on training frequency) to get an initial TDEE figure — use the TDEE calculator or the maintenance calorie calculator to do the arithmetic. Then eat at that estimate for 2–3 weeks, track your weight daily, and use the resulting weight trend to calculate your real TDEE empirically — because the formula is only an average and will be meaningfully off for a large portion of people.

Marcus Chen

NSCA-CPT, MS Exercise Science · Reviewed May 2026

The mechanism

Why this actually matters.

Total daily energy expenditure can vary by ±200–400 kcalbetween two people with identical height, weight, age, and reported activity levels. That gap is not a rounding error — at the low end it represents the difference between maintaining weight and losing nearly half a pound per week without any intentional deficit. The formula treats everyone as an average, but your actual TDEE is shaped by factors it does not measure: your lean muscle mass relative to fat mass, your baseline non-exercise activity thermogenesis (NEAT — the fidgeting, walking, and incidental movement that accounts for 15–30% of total daily calorie burn), the activity of brown adipose tissue, and even gut microbiome composition, which affects how efficiently you extract calories from food. Setting your nutrition targets from a formula alone and never verifying against actual data is how people spend months eating at what they believe is a 500 kcal deficit and seeing no meaningful fat loss — or conversely, cutting aggressively below a number that was already too low.

The method

Five steps to find your real TDEE.

01Start with Mifflin-St Jeor × activity multiplier

The Mifflin-St Jeor equation calculates your basal metabolic rate — the calories your body burns at complete rest. For males: BMR = (10 × weight in kg) + (6.25 × height in cm) − (5 × age) + 5. For females: the same formula minus 161 instead of adding 5. Multiply the resulting BMR by an activity factor: 1.2 for sedentary (desk job, no exercise), 1.375 for lightly active (1–3 sessions per week), 1.55 for moderately active (3–5 sessions), 1.725 for very active (6–7 hard sessions), and 1.9 for extremely active (physical job plus daily training). This gives your estimated TDEE. Use the TDEE calculator or the maintenance calorie calculator to get this number without doing the arithmetic manually. This figure is your starting hypothesis — not your answer.

02Eat at the formula estimate for 2–3 weeks and track weight daily

Log everything you eat at the calorie level your formula produced. Weigh yourself each morning under consistent conditions: immediately after waking, after using the bathroom, before eating or drinking anything. Do this every day for at least 14 days and ideally 21 days. Do not rely on a single weigh-in — daily body weight fluctuates by 1 to 3 lbs due to water retention, sodium intake, glycogen levels, and digestive contents. What you need is a trend, not a data point. Record each day’s weight in a spreadsheet or a tracking app. The goal is to observe whether your weight is genuinely stable at the formula’s calorie level, or whether it is drifting up or down over the full test period.

03Interpret the weight trend

After 2–3 weeks of consistent eating and daily weighing, look at the direction of your weight trend. Three possible outcomes: your weight is stable (±0.5 lbs across the full period) — the formula was accurate for you, and your maintenance is roughly what it predicted. Your weight increased by a meaningful amount — your actual TDEE is lower than the formula estimated; you were in a calorie surplus at the formula’s number. Your weight decreased — your actual TDEE is higher than the formula estimated; you were in a deficit without intending to be. The direction and magnitude of the change is the data you need for step four.

04Calculate your empirical TDEE from the weight change

One pound of body fat stores approximately 3,500 kcal. This number lets you work backwards from a weight change to a calorie discrepancy. Example: you ate at 2,400 kcal for 21 days and lost 1.5 lbs. Total calorie deficit over the period = 1.5 × 3,500 = 5,250 kcal. Daily deficit = 5,250 ÷ 21 = 250 kcal. Your real TDEE = logged calories + daily deficit = 2,400 + 250 = 2,650 kcal, not 2,400. Conversely, if you gained 1.5 lbs eating at 2,400 kcal, your real TDEE is 2,400 − 250 = 2,150 kcal. This empirical figure replaces whatever the formula told you. It is based on what your body actually did, not a population average. Also see the calorie deficit calculator to sanity-check the arithmetic.

05Use your empirical TDEE as your new nutrition baseline

Your empirical TDEE is now the number you use to set any calorie target — deficit for fat loss, surplus for muscle gain, or the maintenance number itself if that is your goal. From this point forward, recalibrate every 4–6 weeks, because TDEE shifts as your weight and body composition change. A 20 lb weight loss will lower your TDEE by 200–300 kcal, depending on how much of that loss was fat versus lean mass. A dedicated macro app that learns your TDEE over time removes the manual recalibration step by treating every logged week as another data point that refines the estimate automatically.

What goes wrong

Common mistakes that
corrupt the test.

The most common error is changing calorie intake mid-test because the number feels too high or too low. If you eat at 2,400 kcal for the first ten days and then drop to 2,100 kcal for the final eleven because you are anxious about gaining, the 3-week weight change now reflects a blended calorie intake, not the 2,400 kcal figure you intended to test. You cannot back-calculate a clean empirical TDEE from inconsistent data. Commit to the test intake for the full period without adjustment. If you find yourself genuinely gaining weight at an uncomfortable rate, your formula estimate was significantly too high — which is exactly what the test is designed to surface. You have your answer faster than expected; you can stop and recalculate.

The second mistake is inaccurate tracking during the test window. The empirical method only works if your logged calorie intake is actually close to your real calorie intake. Estimating portion sizes by eye, forgetting to log cooking oils and condiments, or relying on inaccurate database entries for restaurant meals all introduce error that makes the final calculation meaningless. This is the one period where it is worth using a food scale for your main meals and being genuinely precise. Three weeks of careful tracking gives you a maintenance number you can use for years. Three weeks of approximate tracking gives you a number that is only marginally more reliable than the formula you started with.

Third: weighing at inconsistent times of day or under inconsistent conditions. Body weight can vary by 2–3 lbs across the course of a single day depending on food volume, hydration, and digestive contents. Weighing at 7 am on Monday and at 9 pm on Friday and comparing those two numbers tells you nothing useful about calorie balance. Weigh yourself at the same time each day under the same conditions — first thing in the morning after the bathroom, before eating or drinking — and average the week rather than relying on individual readings.

Fourth: not accounting for water retention during hormonal cycles. For people who menstruate, the luteal phase (the two weeks before menstruation) typically produces 1–3 lbs of water retention that has nothing to do with fat gain. If your 3-week test period spans the late luteal phase, your scale trend will show apparent weight gain that vanishes within a few days of the next cycle. The cleanest test window is the follicular phase (immediately after menstruation ends), when hormonal water retention is at its lowest and weight is most stable. If that is not possible, run the test for a full 4 weeks to capture a complete cycle and average across the entire period rather than reading the last few days in isolation.

Real example

Pat’s maintenance was 200 kcal lower than predicted.

The starting profile

Pat is 33 years old, 178 lbs, moderately active (4 gym sessions per week). Mifflin-St Jeor × 1.55 produces an estimated TDEE of 2,450 kcal. Pat eats at exactly 2,450 kcal for 3 weeks, logs everything accurately with a food scale, and weighs in every morning at the same time.

At the end of 21 days, Pat has gained 0.8 lbs. The calculation: 0.8 lbs × 3,500 kcal = 2,800 kcal total surplus over 3 weeks, or about 133 kcal per day. Real TDEE = 2,450 − 133 ≈ 2,317 kcal. Rounding to the nearest 50 for practical use: 2,300 kcal— not 2,450 kcal. The formula was off by roughly 6%, which is within the typical 15–20% range of error and is consistent with what the research on energy expenditure prediction shows for individual variation.

What this changes

Formula-based deficit target (2,450 − 500)

1,950 kcal

Actual deficit at 1,950 kcal (real TDEE is 2,300)

350 kcal/day

Corrected target for 500 kcal/day deficit (1 lb/week)

1,800 kcal

Without the empirical test, Pat would have been eating at 1,950 kcal believing it was a 500 kcal deficit and losing fat at roughly 0.7 lbs per week instead of 1 lb. Not a disaster, but a 30% shortfall on the intended rate — the kind of gap that compounds over months. Worse, if Pat had used the formula-based 1,950 as the baseline for further adjustments, any subsequent calorie changes would be anchored to a wrong number. The corrected target of 1,800 kcal — based on the empirical TDEE of 2,300 minus 500 — produces the intended 1 lb per week fat loss rate and gives Pat a reliable anchor for all future nutrition decisions.

Zenith tracks your weight trend alongside your logged intake and builds an empirical TDEE estimate automatically over time, replacing formula guesses with data from your own body — try it free.