You can lose fat and gain muscle at the same time — the energy arithmetic allows it, and the trained-population literature reports it — but it is slow, it is invisible on the scale, and it works better for some starting points than others. The reason it is possible at all is that fat and lean tissue are not an even energy swap: a kilogram of fat stores several times the energy of a kilogram of lean tissue. This page does the arithmetic, says who it tends to work for and why, shows what the trials measured, gives the calories, and shows how to tell whether it is happening. The trials report group averages, not a promise for any one person.
Why it is possible on paper
The planning values we use here come from Hall's 2008 analysis of the energy deficit per unit of weight loss, which puts the energy density of lean body mass at 7.6 MJ/kg and of body fat at 39.5 MJ/kg1 — about 1,800 kcal per kilogram of lean tissue and about 9,400 kcal per kilogram of fat tissue. (Elsewhere on this site weight change is planned at 7,700 kcal/kg, a blended figure for the mixed tissue a typical diet loses; the split matters here because the two tissues are moving in opposite directions.)
Suppose over ten weeks someone gains 1 kg of lean tissue and loses 1 kg of fat, and the scale does not move at all:
energy stored in new lean tissue = 1 × 1,800 = 1,800 kcalenergy released from fat = 1 × 9,400 = 9,400 kcalnet = 9,400 − 1,800 = 7,600 kcal deficit over 70 days ≈ 109 kcal per day
That person was in a genuine 109 kcal daily deficit for ten weeks with nothing to show for it on the scale. Double the swap — 2 kg each way, which would be a very good ten weeks — and it is a 217 kcal daily deficit at constant weight.
| Lean gained / fat lost over 10 weeks | Net energy change | Per day | Scale change |
|---|---|---|---|
0.5 kg / 0.5 kg | −3,800 kcal | −54 kcal | 0 |
1 kg / 1 kg | −7,600 kcal | −109 kcal | 0 |
2 kg / 2 kg | −15,200 kcal | −217 kcal | 0 |
1 kg / 2 kg | −17,000 kcal | −243 kcal | −1 kg |
Three things follow. Recomposition needs only a small energy deficit, or none, at the scale level. It needs a reason for the body to build lean tissue — that is the training, not the food. And it is a slow process measured in kilograms per quarter, not per week.
In plain terms: because fat is energy-dense and muscle is not, you can be losing one and building the other while the scale sits still.
Who it tends to work for
The 2020 review by Barakat and colleagues in the *Strength and Conditioning Journal* asked exactly the question in this page's title for trained individuals, and concluded that they can — naming progressive resistance training together with supportive nutrition as the critical factors, and noting gaps in the measurement methods used to detect it2. We cite it for that scope and those conditions; it is a narrative review, not a trial with an effect size we can hand you.
Beyond the review, the arithmetic itself says who has the easiest time:
| Starting point | Why the arithmetic is kinder | Why it is harder |
|---|---|---|
| New to lifting, or returning after a long break | The fastest lean-tissue gains are at the start; the 1,800 kcal/kg side of the ledger fills quickly | — |
| Carrying more body fat | More stored energy to draw the 9,400 kcal/kg side from without a large intake deficit | — |
| Lean and well trained | — | Lean gains are slow, so the deficit side must be small too; progress is measured in hundreds of grams |
| Training without progression | — | Nothing asks the body to build; the arithmetic has no lean side |
None of that is a health claim; it is where the two tissue budgets have room to move. A well-trained lean lifter is not excluded, but for them a deliberate small surplus or small deficit — a slow bulk or slow cut — usually produces a clearer result than trying to hold both.
What the trials measured
The clearest proof of principle is a 2016 trial3. Forty young men who carried extra body fat and did not lift were given every meal for four weeks, at about 40% below their estimated energy needs. They trained six days a week under supervision: full-body lifting circuits, sprint intervals, a time trial and a body-weight circuit. Half ate 2.4 g of protein per kg a day, half 1.2 g3. Body composition was measured with a four-compartment model, which combines body volume, body water and bone for a more exact split than a scan alone.
On average, the higher-protein group gained 1.2 kg of lean body mass while the lower-protein group held steady (+0.1 kg); both lost fat, the higher-protein group more (4.8 vs 3.5 kg)3. Scale weight changed the same in both groups, and strength and fitness improved alike3. Lean mass here means everything that is not fat, so it includes water, not only muscle.
Four weeks at about 40% below energy needs, by protein intake · kg
2.4 g/kg/day protein 1.2 g/kg/day protein
Trained people have been tested too. In a 2026 trial, 30 adults with at least a year of lifting trained four days a week for ten weeks5. Two groups ate 2.5 g/kg of protein a day, one at estimated maintenance and one 250 kcal below it; a control group trained without diet guidance. Fat-free mass rose about the same in both diet groups (+0.97 and +1.04 kg on average, DXA). Fat mass fell 2.94 kg in the deficit group and 1.41 kg at maintenance, the second not statistically significant. The control group barely changed5.
Ten weeks of lifting, trained adults: fat mass vs fat-free mass change · kg
Fat mass Fat-free mass
The size of the deficit matters. A 2022 meta-analysis of randomised trials found lean-mass gains from resistance training were smaller in an energy deficit than without one, while strength gains were similar; its meta-regression put the deficit that prevented lean-mass gains at about 500 kcal a day4. That fits the arithmetic above: recomposition lives close to maintenance.
In plain terms: in controlled trials, lean mass went up while fat went down, most clearly in beginners with fat to lose, with plenty of protein and hard training, and with the deficit small or the setting tightly supervised.
Calories for recomposition — the maintenance band
Recomposition lives in a narrow band around maintenance. The calculator gives you three settings for it. Take an 80 kg, 178 cm, 30-year-old male profile training 4–6 times a week with moderate daily activity: BMR 1,768 kcal, TDEE 2,740 kcal.
| Setting | Rate | Weekly change planned | Daily target | Best suited to |
|---|---|---|---|---|
| Slow cut | lose, conservative 0.3% | −0.24 kg | 2,476 kcal | More fat to lose than muscle to gain |
| Maintain | maintain | 0 | 2,740 kcal | The classic recomposition setting |
| Slow gain | gain, conservative 0.3% | +0.24 kg | 3,004 kcal | Lean, prioritising muscle, willing to see the scale rise a little |
The difference between rows is 264 kcal (80 × 0.003 × 1,100) either side of TDEE. For comparison, the 2026 trial's deficit group ate 250 kcal below estimated maintenance5, close to the slow-cut row, and the meta-regression's 500 kcal point sits well beyond it4. Protein is set from body weight, not from the calorie total — 1.6 g/kg on the moderate tier, 128 g here, with the 0.2 g/kg bump only on the lose goal — so it barely changes across the rows. The macro calculator shows the full split for each.
A practical band to hold the trend inside: if the trend weight drifts more than about 0.7 kg from where you meant to be, in either direction, for two weeks, the setting is not maintenance for you — move one row. Inside that band, the scale is telling you nothing useful and the tape takes over. (This is a tracking convention, not a rule from the calculator; the 35% cap and the floors still apply to the lose row as on every loss goal.)
How to tell it is happening (trend + tape)
The scale is the wrong instrument, so use three others together:
- Trend weight, flat or nearly. A trend moving less than the row above planned is consistent with the table's first three rows — and with nothing happening at all. On its own it proves nothing.
- Tape. A waist measurement taken the same way, same morning, every week or two. The tape measures a size the scale cannot see; a centimetre or two off the waist with a flat trend is the first real signal.
- Training log. Rising estimated 1RMs or more reps at the same load across the block. Strength is not a direct measure of muscle, but a flat trend with a shrinking waist and rising e1RMs over four to eight weeks has very few other explanations. The 1RM calculator turns any logged set into a comparable number.
Give it a month before judging, and eight weeks before changing rows on the strength of the tape alone. Water and glycogen move the scale by more in a day than recomposition moves it in a month — why the scale jumps overnight has the numbers.
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Hall, *Int J Obes*1 | Analysis of the energy stored in body tissues | Lean tissue about 7.6 MJ/kg, fat about 39.5 MJ/kg: an uneven swap | 2008 |
| Longland et al., *Am J Clin Nutr*3 | 40 young men new to lifting, extra body fat; 4 weeks, ~40% deficit, all meals provided | 2.4 g/kg protein: lean mass +1.2 kg, fat −4.8 kg; 1.2 g/kg: lean mass +0.1 kg, fat −3.5 kg (group averages) | 2016 |
| Barakat et al., *Strength Cond J*2 | Narrative review of recomposition in trained people | Trained people can do it with progressive training and supportive nutrition | 2020 |
| Murphy & Koehler, *Scand J Med Sci Sports*4 | Meta-analysis of randomised training trials of at least 3 weeks | Deficits blunted lean-mass gains but not strength; about 500 kcal a day prevented lean-mass gains | 2022 |
| Vargas-Molina et al., *Eur J Appl Physiol*5 | 30 adults with 1+ year of lifting; 10 weeks, 2.5 g/kg protein at maintenance or −250 kcal | Fat-free mass about +1 kg in both diet groups; fat −2.94 kg with the deficit | 2026 |
Limits worth knowing
- Short and small. Four and ten weeks, 40 and 30 people35.
- Mostly men. The 2016 trial was all men; the 2026 trial had 4 women among 30.
- Beginners with fat to lose. The clearest result was in untrained young men with extra body fat, under an extreme, supervised regime3.
- Two things changed at once. In the 2016 trial the higher-protein diet was also lower in fat, so the effect cannot be pinned on protein alone3.
- Fat-free mass is not muscle. It includes water, organs and bone; DXA and four-compartment models estimate it, they do not image muscle5.
- Abstract only for the 2022 meta-analysis; the two trials were read in full.
Latest research
The newest trial we read, published in July 2026, is the ten-week study in trained adults above5. Its authors note that the capacity for recomposition shrinks as training status and leanness rise, that their participants still averaged about 25% body fat, and that leaner, more experienced lifters may see less5. It adds a controlled comparison of maintenance and a small deficit at the same high protein intake; it does not show that everyone can recompose.
The short version
Yes, on paper and in the trained-population literature2. Fat stores about 9,400 kcal/kg and lean tissue about 1,8001, so swapping a kilogram each way is a 7,600 kcal deficit at constant weight — about 109 kcal a day over ten weeks. It works best for people new or returning to lifting and those with more fat to draw on; it needs progressive training to fill the lean side. Calories sit at maintenance or ±0.3% of body weight per week; judge it by the tape and the training log, because the scale is designed not to notice.
- In a 4-week trial of beginners, the higher-protein group gained lean mass while losing fat, with the same change on the scale3.
- Trained adults gained about
1 kgof fat-free mass on average at maintenance or250 kcalbelow it5. - Bigger deficits blunted lean-mass gains; about
500 kcala day stopped them on average4.
Educational overview only — not medical advice. Zyra is a training and nutrition tracking tool; the rates, caps and floors on this page describe how the calculator behaves, not what any individual should do.
