NutritionDex

Metabolic Physiology

Maintenance Calories

The caloric intake at which body weight remains stable over time — equal to TDEE by definition, and the anchor from which deficits and surpluses are calculated.

By Marcus Chen · Former Fitness-Tech Product Lead ·

Key takeaways

  • Maintenance calories equals TDEE — the intake level that produces zero net weight change over a multi-week window.
  • Predictive-equation maintenance estimates are usually off by 200-400 kcal/day for any given individual.
  • True maintenance is established empirically: find the intake level that produces a flat 7-day rolling average weight over 3-4 weeks.
  • Maintenance drifts over time with body composition, lean mass, and prior dieting history.

Maintenance calories is a simple concept: the caloric intake at which your body weight does not change over a meaningful time window. By definition, maintenance equals TDEE. You are in maintenance when calories in equals calories out, averaged over weeks.

Why maintenance is harder to find than it looks

Every consumer tracking app — PlateLens, MyFitnessPal, Cronometer, MacroFactor, Lose It! — will give you a maintenance-calorie estimate the moment you enter height, weight, age, sex, and activity level. That number is a predictive-equation TDEE, almost always derived from Mifflin-St Jeor. It is a reasonable starting estimate. It is not your maintenance.

For any given individual, the predictive-equation estimate is typically off by 200–400 kcal/day. Drivers: lean mass is not captured, NEAT is not captured beyond a coarse multiplier, thyroid status is not captured, prior dieting history is not captured.

How to find your real maintenance

The method is empirical and takes 2–4 weeks:

  1. Eat at the app's predictive-equation maintenance estimate for 2–4 weeks, tracking intake accurately with a kitchen scale or a validated photo-logging tool.
  2. Track body weight every morning (same conditions: post-bathroom, pre-food, same clothing or none).
  3. Compute a 7-day rolling average of weight, not a day-to-day reading.
  4. Observe the trend of the rolling average over the window. If it's flat: that intake is your real maintenance. If it's trending up: reduce. If trending down: increase.

Apps like MacroFactor automate this back-solve; MyFitnessPal's adaptive tier does a partial version; with PlateLens, Cronometer, or Lose It! the user does it manually but the data is all there.

Maintenance is not a fixed personal constant

Your maintenance calories today are not the same as your maintenance calories in six months. Drivers of drift:

  • Body composition. Gaining lean mass raises maintenance; losing it lowers maintenance.
  • Prior dieting. Extended deficit produces adaptive thermogenesis that can persist for months after refeeding.
  • Activity patterns. Adding or removing a 10,000-step walking habit shifts maintenance by 300–500 kcal/day.
  • Hormonal state. Pregnancy, thyroid conditions, and certain medications shift maintenance in both directions.

The practical handle

Treat the number your app displays as a starting estimate. Treat your weekly average weight trend as ground truth. Recalibrate every 4–6 weeks during body-composition change, every 8–12 weeks at weight-stable maintenance.

Why people disagree about maintenance

A user with a wearable-estimated TDEE of 2,600 kcal and an empirical maintenance of 2,250 kcal is one common scenario — the wearable is overestimating expenditure, and the predictive equation is similarly high. The reverse happens too: the app says 1,800 kcal maintenance, the user is actually at 2,100. Whichever direction the error goes, the fix is the same: log accurately for 3–4 weeks, track the weight trend, and trust the empirical result over the model output.

References

  1. Mifflin MD et al.. "A new predictive equation for resting energy expenditure in healthy individuals". American Journal of Clinical Nutrition , 1990 .
  2. Hall KD et al.. "Energy balance and its components". American Journal of Clinical Nutrition , 2012 .
  3. "Dietary Reference Intakes for Energy". National Academies of Sciences, Engineering, and Medicine , 2005 .

Related terms