Yes, measurably — and then most of the fat goes back where it came from#
Caffeine raises the calories you burn. This part is not in dispute and has been measured with good instruments for decades: in a double-blind trial giving healthy adults placebo, 100, 200 and 400 mg, energy expenditure rose dose-dependently, and it did so in people who already drank coffee habitually2. The bump is real, it scales with dose, and it does not require you to be caffeine-naive.
What almost nobody reports is the second half of the measurement. When eight healthy men were given 10 mg/kg of caffeine — roughly 700 mg for a 70 kg adult, about seven strong coffees at once — and traced with labelled palmitate for four hours, energy expenditure rose 13.3 ± 2.2 percent and lipid turnover doubled. But the proportion of mobilised fat that was actually oxidised fell, from 34 percent before the dose to 24 percent after it; the other 76 percent was recycled straight back into storage1. Caffeine got fat moving. It did not get much of it burned. This article is about the gap between those two sentences, and it is the gap the whole thermogenic-supplement shelf is built in. The general case against shopping for foods that raise your burn is in do 'metabolism-boosting' foods work; this is the strongest single member of that category, examined closely.
What 13 percent is worth in calories#
Before going further, put a number on the headline. A 13 percent rise over four hours, applied to a resting rate near 1,500 kcal/day, is about 32 extra calories (our arithmetic on Acheson's percentage and a representative resting rate — not a figure the paper reports). That is the return on seven coffees' worth of caffeine, taken at once, in young men, under laboratory conditions, at the upper end of any dose anyone would recommend.
The honest way to hold the finding is that caffeine is the strongest genuine thermogenic in ordinary food and drink, and that being the strongest member of this category is not an impressive credential. The thermic effect of the meal you drank it with is larger and more reliable — see the thermic effect of food — and it is a percentage of intake, so it can never net you a surplus either.
The heat is coming from cycling, not from fat disappearing#
The more interesting question is where the extra heat is generated, because the answer contradicts the marketing directly.
Acheson's tracer data shows the shape. Oxidative disposal of free fatty acids rose 44 percent, from 236 ± 21 to 340 ± 16 µmol/min — a genuine increase in fat being burned. Non-oxidative disposal rose 2.3-fold over the same window, from 455 ± 66 to 1,054 ± 242 µmol/min1. Fat came out of storage far faster than it was consumed, so the surplus went back in. The authors' framing is that lipid mobilisation alone is insufficient to drive lipid oxidation, and that large increments in turnover produce small increments in oxidation. Breaking down a triglyceride and re-esterifying it is not free — the cycle itself costs energy — which is a plausible source of some of the heat.
The earlier dose-response work points the same way from a different angle. The size of the thermogenic response correlated with plasma lactate (r = 0.79, p < 0.000001) and plasma triglyceride (r = 0.53, p < 0.02), and the authors concluded that lactate and triglyceride production together with increased vascular smooth-muscle tone "may be responsible for the major part of the thermogenic effect of caffeine"2.
Read the two studies together and the mechanism looks less like a furnace and more like a busier engine idling. Substrate is being shuttled around, blood vessels are working harder, and the metabolic cost of all that activity shows up as heat. It is real expenditure. It is just not the mechanism the phrase "fat burner" is selling, and the difference matters when someone concludes that a stimulant is emptying their fat cells.
Tolerance: what is proven, and what is assumed#
The standard next move is to say the effect disappears once you are used to caffeine. That claim is repeated far more confidently than the evidence supports, and the confusion is specific enough to name.
What has been demonstrated cleanly is cardiovascular and hormonal tolerance. When caffeine-naive people were dosed repeatedly, the increases in blood pressure, heart rate, plasma epinephrine, norepinephrine and renin activity showed near-complete tolerance within the first one to four days, with no long-term effect on any of those variables3. That is a strong, well-replicated result — about the pressor response.
Thermogenesis is a different endpoint, and it has not been shown to habituate the same way. The clearest counter-evidence is that the dose-response trial above was run in subjects with moderate habitual caffeine consumption and still found expenditure rising with dose2. So the fair statement is narrower than either camp's version: your blood pressure stops responding within days; whether your metabolic rate does is not established, and the studies that report a thermogenic effect in habituated drinkers are evidence it does not vanish entirely.
This matters less than it sounds, because 30-odd calories is not worth defending in either direction. It matters as a caution about how a well-evidenced finding in one system gets exported to another and repeated as fact.
During exercise, and only above a threshold#
The one context where caffeine's effect on fat is more than a rounding error is training — with two conditions attached that are almost never quoted alongside the headline.
Pooling 19 crossover trials from 1978 to 2020, acute caffeine raised the fat-oxidation rate during submaximal aerobic exercise (SMD 0.73, 95% CI 0.19 to 1.27, p = 0.008), lowered the respiratory exchange ratio (SMD −0.33) and raised oxygen uptake slightly (SMD 0.23). Doses above 3.0 mg/kg were required; at or below 3.0 mg/kg there was no effect at all (p = 0.78). And the effect reached significance only in sedentary and untrained participants, not in trained or recreational athletes4.
| Condition | What the pooled data shows |
|---|---|
| Dose above 3 mg/kg (≈210 mg at 70 kg) | Fat oxidation rises, SMD 0.73 |
| Dose at or below 3 mg/kg | No detectable effect (p = 0.78) |
| Untrained or sedentary participants | Effect significant |
| Trained or recreational athletes | Effect not significant |
There is a quiet irony in that last pair of rows. The people for whom pre-workout caffeine measurably shifts fuel use are the ones least likely to be buying pre-workout, and the trained athletes who consume the most of it are the group in whom the fat-oxidation effect did not reach significance. Note also what the outcome is: which fuel is being used, not how much total energy is spent. Oxidising more fat during a ride does not by itself create a deficit.
Does any of it reach the scale?#
A dose-response meta-analysis of 13 randomized trials in 606 participants found that increasing caffeine intake was associated with progressive reductions in weight, BMI and fat mass — pooled betas of 0.29 (95% CI 0.19 to 0.40), 0.23 (0.09 to 0.36) and 0.36 (0.24 to 0.48), which the authors translate as a 22, 17 and 28 percent greater reduction respectively for every doubling of dose5.
That sounds substantial until you look at what is missing and what is present. What is missing is an absolute anchor: a 22 percent larger reduction is meaningless without knowing 22 percent of what, and a proportional coefficient can describe a doubling of a very small number. What is present is heterogeneity of I² = 91.2, 93.0 and 94.0 percent across the three outcomes — a formal statement that the trials scatter so widely that the pooled estimate poorly describes any of them. The authors' own verb is careful: caffeine intake "might promote" reduction. Set against a measured 30-odd calories a day, that is the right level of confidence.
One disclosure standard, applied in both directions. The two syntheses in this article that report null or bounded effects — the fat-oxidation meta-analysis and the sleep meta-analysis below — both declare no conflicts of interest. In the other direction, a widely cited trial reporting an approximately 8 percent rise in resting energy expenditure sustained for four hours after a commercial "thermogenic" supplement was funded by the supplement's manufacturer, and two of its authors were employees of that company7. The study is competently run and the disclosure is properly made. It is still a manufacturer measuring its own product, and this category is unusually full of them.
The bill that arrives at bedtime#
The strongest argument against chasing caffeine's thermogenic effect is not that the effect is small. It is that the same molecule has a much larger and better-quantified effect in the opposite direction.
Pooling 24 studies, caffeine reduced total sleep time by 45 minutes and sleep efficiency by 7 percent, increased sleep-onset latency by 9 minutes and wake-after-sleep-onset by 12, added 6.1 minutes of light N1 sleep and removed 11.4 minutes of deep N3 and N4 sleep. To avoid losing sleep time the authors calculate that a 250 mL coffee at 107 mg should be taken at least 8.8 hours before bed, and a 217.5 mg pre-workout serving at least 13.2 hours before6.
Forty-five fewer minutes of sleep, with the deep portion disproportionately cut, is a far bigger lever on next-day appetite and food choice than 30 calories of substrate cycling is on your energy balance — and it runs the wrong way. That trade is worked through in improving sleep for weight loss. The practical reading is unromantic: drink coffee because you like it and because it helps you train, keep it to the first half of your day, and do not enter a single calorie of it on the credit side of your ledger. The organ-level reason no beverage can move this number much is in metabolism explained, and the same arithmetic disposes of the green tea claim.
FAQ#
Is caffeine actually burning fat, or just moving it around?#
Mostly moving it around. With labelled tracers, 10 mg/kg of caffeine doubled lipid turnover and raised oxidative disposal of free fatty acids by 44 percent — but non-oxidative disposal rose 2.3-fold, so the share of mobilised fat that was actually oxidised fell from 34 percent to 24 percent. Fat left storage faster than it was consumed, and three-quarters of it went back. The increase in energy expenditure is real; the fat loss it implies is largely not.
How much caffeine would I need to burn a meaningful number of calories?#
More than is sensible, for a return that still is not meaningful. The clearest thermogenic measurement used 10 mg/kg — roughly 700 mg for a 70 kg adult, about seven strong coffees — and produced a 13.3 percent rise in expenditure over four hours, which on a typical resting rate works out near 30 calories. Doses that shift fat oxidation during exercise start above 3 mg/kg. There is no dose at which the burn side becomes a weight-loss strategy.
Does caffeine still work if I drink coffee every day?#
For the metabolic effect, probably yes, though the evidence is thinner than the confident claims on either side. Near-complete tolerance to caffeine's effects on blood pressure, heart rate and catecholamines develops within one to four days of regular use. Thermogenesis is a separate endpoint and has not been shown to habituate the same way — the classic dose-response trial was conducted in habitual moderate coffee drinkers and still found expenditure rising with dose.
Sources#
- Acheson KJ, Gremaud G, Meirim I, et al. Metabolic effects of caffeine in humans: lipid oxidation or futile cycling? Am J Clin Nutr. 2004;79(1):40-46.
- Astrup A, Toubro S, Cannon S, Hein P, Breum L, Madsen J. Caffeine: a double-blind, placebo-controlled study of its thermogenic, metabolic, and cardiovascular effects in healthy volunteers. Am J Clin Nutr. 1990;51(5):759-767.
- Robertson D, Wade D, Workman R, Woosley RL, Oates JA. Tolerance to the humoral and hemodynamic effects of caffeine in man. J Clin Invest. 1981;67(4):1111-1117.
- Collado-Mateo D, Lavín-Pérez AM, Merellano-Navarro E, Del Coso J. Effect of acute caffeine intake on the fat oxidation rate during exercise: a systematic review and meta-analysis. Nutrients. 2020;12(12):3603.
- Tabrizi R, Saneei P, Lankarani KB, et al. The effects of caffeine intake on weight loss: a systematic review and dose-response meta-analysis of randomized controlled trials. Crit Rev Food Sci Nutr. 2019;59(16):2688-2696.
- Gardiner C, Weakley J, Burke LM, et al. The effect of caffeine on subsequent sleep: a systematic review and meta-analysis. Sleep Med Rev. 2023;69:101764.
- Outlaw J, Wilborn C, Smith A, et al. Effects of ingestion of a commercially available thermogenic dietary supplement on resting energy expenditure, mood state and cardiovascular measures. J Int Soc Sports Nutr. 2013;10:25.
