If you drank 10 cups of filter coffee a day, you would take in roughly 600 to 1,200 mg of caffeine, with about 900 mg as a fair middle estimate. That is more than twice the 400 mg a day that EFSA, the US FDA and the American Heart Association treat as safe for healthy adults. Spread across a day it is very unlikely to kill you, but it would wreck your sleep, push your heart rate and anxiety up, and keep about 300 mg of caffeine circulating at bedtime.
- ~900 mg of caffeine in ten tumblers, at an estimated 90 mg each.
- 400 mg a day is the limit EFSA, the US FDA and the AHA accept for healthy adults.
- About 300 mg would still be in you at 11 pm, with a 5-hour half-life.
- 3–7 hours is the adult half-life; smoking, the pill, pregnancy and genes move it a lot.
- About 10 g is the estimated lethal dose: roughly 100 cups, almost always reached with pills or powder.
Filter kaapi is the drink most of South India runs on: a dark decoction dripped through a brass or steel filter, cut with hot milk and sugar, and served in a tumbler and davara. Ten a day is not a hypothetical in some offices and newsrooms.
So the question is a fair one: what does caffeine actually do at that dose, how long does it stay in you, and how far is 10 cups from the dose that puts people in hospital? The answer needs three numbers:
- how much caffeine is in a tumbler;
- how fast your liver clears it;
- where the toxic thresholds sit.
How much caffeine is in one cup of filter coffee?
A cup of filter coffee carries about 90 mg of caffeine by the European Food Information Council's standard figure for a 200 ml cup, and South Indian filter coffee lands in a similar band, roughly 50 to 150 mg depending on the powder, the chicory share and how strong the decoction is. Nobody has published a lab measurement of a typical tumbler of kaapi, so treat 90 mg as an estimate, not a fact.
Three things move the number
- The bean. Robusta generally carries more caffeine than arabica, and South Indian blends often lean on robusta grown in Karnataka and Kerala.
- Chicory. FSSAI rules require a coffee-chicory mixture to contain at least 51% coffee by mass, and from 1 July 2026 the coffee and chicory percentages must be printed on the front of the pack. Chicory root has no caffeine, so an 80:20 blend has about a fifth less caffeine than pure coffee of the same weight.
- Extraction. A slow drip through a fine powder pulls most of the caffeine out, because caffeine dissolves readily in hot water.
The American Heart Association's 2026 statement puts regular brewed coffee at 9.4 to 20.6 mg of caffeine per fluid ounce, which is about 32 to 70 mg per 100 ml. A tumbler of kaapi is usually less than 150 ml in total, but the decoction part is strong and the milk part has no caffeine, so the two effects partly cancel.
| Drink or product | Serving | Caffeine | Source |
|---|---|---|---|
| Filter coffee | 1 cup, 200 ml | about 90 mg | EUFIC, from EFSA data |
| South Indian filter kaapi | 1 tumbler | 50–150 mg (estimate) | Author's estimate from the rows above and below |
| Brewed coffee | 8 fl oz, 237 ml | 75–165 mg | AHA 2026 (9.4–20.6 mg per fl oz) |
| Espresso | 60 ml | about 80 mg | EUFIC |
| Tea | 220 ml | about 50 mg | EUFIC |
| Cola | 355 ml can | about 40 mg | EUFIC |
| Energy drink | 250 ml can | about 80 mg | EUFIC |
| Energy shot | per fl oz | 40–69 mg | AHA 2026 |
| Pure caffeine powder | less than ½ teaspoon | about 1,200 mg | US FDA |

What does caffeine do inside your body?
Caffeine works mainly by blocking adenosine receptors in the brain, which is why it keeps you awake: adenosine builds up during waking hours and signals sleep pressure, and caffeine sits in its receptors without activating them. The molecule looks enough like adenosine to fit the lock.
Yamamoto and colleagues add that at high doses it also raises calcium inside cells, releases noradrenaline and sensitises dopamine receptors, and those are the routes to a racing heart and arrhythmia.
How fast it arrives and leaves
Absorption is fast and nearly complete. Effects begin within about 15 minutes of drinking, and blood levels peak within about an hour. The liver then does almost all the clearing, through an enzyme called CYP1A2 that strips methyl groups off the molecule. How fast that enzyme works decides how long each cup stays with you.
The novelist Honoré de Balzac left one of the most detailed first-person records of chasing a caffeine high. In the 1830s he wrote an essay on modern stimulants as an appendix to Brillat-Savarin's Physiologie du goût, and its coffee section reads like a tolerance diary.
He noted that coffee's power fades with habit, and described escalating to keep it: crushing the beans finer, using less water and cold infusion, then doubling or tripling cups. The last stage, which he said was for only the most vigorous, was finely pulverised coffee taken dry on an empty stomach. He reported racing thoughts, but also sweating, irritability, bouts of heavy drowsiness and lasting stomach trouble, and said he stopped before it killed him. His remedy was a diet of milk dishes and white meat.
Balzac: The Pleasures and Pains of Coffee, trans. Robert Onopa, Michigan Quarterly Review 35(2), 1996Balzac's escalation is what pharmacology now calls tolerance: regular drinkers adapt, so the same cup feels weaker, while the caffeine load in the body keeps climbing. The next section puts numbers on that load.
A review of human and animal evidence on caffeine's effects on the heart, brain, bones, pregnancy and behaviour, with special attention to children and adolescents. It puts the adult half-life at about 3 to 7 hours, notes that pregnancy can lengthen it by 8 or more hours and oral contraceptives roughly double it, while smoking roughly doubles clearance. CYP1A2, the main enzyme, has at least 150 known gene variants. It accepts about 400 mg a day as the threshold for healthy adults and 100 mg a day for adolescents, and links most serious harm to doses above about 3 mg per kg, usually from pills or stimulant products. The limit: it is a narrative review, so it summarises rather than pools data, and much of the high-dose evidence is from case reports.
Read the paper (free)How much caffeine is still in you at bedtime?
With a 5-hour half-life and a tumbler every 90 minutes from 7 am, ten cups leave about 300 mg of caffeine in your body at 11 pm, which is like downing three cups just before bed. Three cups spread over the morning leave under 50 mg by the same hour.
The working is a simple one-compartment model. Each dose \(D\) decays by half every half-life \(t_{1/2}\), so a single cup taken at time zero leaves \(D \cdot 0.5^{t/t_{1/2}}\) at time \(t\). Ten cups spaced \(\Delta\) hours apart add up as a geometric series:
- Inputs. \(D\) = 90 mg, \(n\) = 10, \(\Delta\) = 1.5 h, \(t_{1/2}\) = 5 h and \(t\) = 16 h (7 am to 11 pm).
- Decay factor. \(90 \times 0.5^{3.2} = 9.8\) mg.
- Series factor. \((2^{3} - 1)/(2^{0.3} - 1) = 7/0.231 = 30.3\).
- Bedtime load. Multiply: about 297 mg.
The model assumes each cup is absorbed instantly, which slightly overstates the early peaks, but the bedtime number is robust because absorption is over within an hour.
See the data as a table
| 3 tumblers (7, 10 am, 1 pm) (mg) | 10 tumblers (every 90 min from 7 am) (mg) | |
|---|---|---|
| 7 am | 90 | 90 |
| 9 am | 68 | 152 |
| 11 am | 130 | 194 |
| 1 pm | 189 | 310 |
| 3 pm | 143 | 319 |
| 5 pm | 108 | 320 |
| 7 pm | 82 | 406 |
| 9 pm | 62 | 391 |
| 11 pm | 47 | 297 |
| 1 am | 36 | 225 |
| 3 am | 27 | 170 |
| 5 am | 21 | 129 |
| 7 am | 16 | 98 |
Source: Author's one-compartment model: 90 mg per tumbler, 5-hour half-life (mid-range of Temple et al., 2017) · Values sampled every two hours, just before any cup due at that time. Instant absorption assumed.
Sensitivity: the half-life
Change the half-life and the story changes a lot. From the same ten cups:
| Drinker | Half-life | Caffeine at 11 pm |
|---|---|---|
| Smoker (clears twice as fast) | 2.5 h | about 130 mg |
| Typical adult | 5 h | about 300 mg |
| Woman on the combined pill (half-life doubled) | 10 h | about 500 mg |
That one variable, set by genes and hormones, is why one person sleeps after an evening kaapi and another lies awake. Use this to check what you just read, with your own numbers:
Your caffeine day
Is 10 cups of coffee a day dangerous?
Ten cups is a dose most healthy adults survive but should not choose: at about 900 mg it is more than twice the 400 mg daily safety threshold, though it is still about a tenth of the dose that kills. The danger depends on how fast it goes in, because the body only ever holds the part it has not yet cleared.
What the regulators say
- EFSA (2015): habitual intake up to 400 mg a day, and single doses up to 200 mg (about 3 mg per kg for a 70 kg adult), raise no safety concerns for healthy non-pregnant adults.
- US FDA: the same 400 mg line, with a warning that toxic effects such as seizures can appear with rapid intake of around 1,200 mg.
- AHA (2026): its scientific statement in Circulation calls up to 400 mg a day generally safe for most adults, roughly 3 to 5 cups of regular black coffee.
| Threshold | Dose | What it means | Source |
|---|---|---|---|
| Single dose with no concern | 200 mg (about 3 mg/kg) | Two average cups at once | EFSA 2015 |
| Daily intake with no concern | 400 mg | About 4 to 5 tumblers | EFSA 2015, FDA, AHA 2026 |
| Pregnancy limit | 200 mg a day | About 2 tumblers | EFSA 2015 |
| Ten tumblers | about 900 mg | Spread over 13 hours, peak body load about 420 mg | This post's model |
| Seizure risk when taken fast | about 1,200 mg | Under ½ teaspoon of pure powder | US FDA |
| Fatal blood level | 80–100 mg per litre | Usually needs 50–100 tablets of 100 mg | Yamamoto et al., 2015 |
| Estimated lethal intake | about 10 g | About 100 cups at once | Yamamoto et al., 2015 |
The reason ten tumblers rarely put anyone in hospital is pharmacokinetics. Drunk 90 minutes apart, the body load in the model never passes about 420 mg, because the liver keeps clearing while you keep drinking. The same 900 mg swallowed in a few minutes, as pills or a pre-workout scoop, is a different event.
You simply cannot drink coffee fast enough to reach 10 g. Almost every caffeine death in the medical literature involves tablets, powders or energy products.
The authors describe the death of an 18-year-old woman after a caffeine tablet overdose, with peripheral blood caffeine of 290 mg per litre, and review the toxicology. They cite oral LD50 values of 200 to 400 mg per kg in rats and 185 mg per kg in mice, an estimated lethal human intake of about 10 g (roughly 100 cups of coffee), and fatal blood concentrations of 80 to 100 mg per litre, which would take 50 to 100 tablets of 100 mg. They name ventricular arrhythmia as the usual mechanism of death, with seizures as another route. The limit: a single case plus a narrative review, and lethal thresholds vary widely between individuals.
Read the paper (free)
What would 10 cups do to your heart, sleep and stomach?
At about 900 mg a day, expect a faster heart rate, jitteriness, anxiety, broken sleep and acid trouble; the heart findings matter most because the AHA now links more than four cups a day with a possible rise in heart failure risk.
- Heart. The AHA's July 2026 statement found that 2 to 4 cups a day was associated with lower risks of heart disease, heart failure and stroke, and randomised trials linked coffee caffeine with lower atrial fibrillation. Above four cups a day the association turned the other way for heart failure. These are observational links, not proof, but ten cups sits well past the point where the benefit curve bends.
- Sleep. With about 300 mg still in the body at bedtime, adenosine cannot do its job. Temple's review states plainly that caffeine disrupts sleep in every population studied, and the next morning's tiredness pushes the first cup earlier. That loop is how ten cups becomes a habit. The mindblog post on sleep and memory consolidation explains what poor sleep costs.
- Anxiety and tremor. Temple's review cites anxiety reported across intakes from 200 to 2,000 mg a day, and DSM criteria for caffeine intoxication usually involve more than 250 mg taken recently: restlessness, nervousness, flushing, a racing heart and stomach upset.
- Stomach. Caffeine and coffee acids irritate the stomach lining in some people; Temple notes gastric irritation, sometimes to the point of vomiting, at high doses. Dark roasts are not automatically gentler: Starbucks Italian Roast, a very dark roast, gave me acidity.
EFSA's panel reviewed human studies on caffeine from all sources, looking at the heart, the nervous system, sleep, anxiety, pregnancy outcomes and interactions with alcohol and exercise. It concluded that single doses up to 200 mg and habitual intake up to 400 mg a day do not raise safety concerns for healthy, non-pregnant adults, and that up to 200 mg a day in pregnancy does not raise concerns for the fetus. It found the data too thin to set a safe level for children and adolescents. The limit: the opinion sets a "no concern" level, not a danger line, and it does not model what happens at 900 mg; it simply cannot vouch for it.
Read the opinion (free)Why do some people handle 10 cups better than others?
Habitual and occasional drinkers respond differently, Temple's review notes, but habit does not change the arithmetic of clearance: the bigger differences in how long caffeine stays come from metabolism.
- Genes. CYP1A2 variants make some people fast metabolisers and others slow ones. A slow metaboliser on ten cups behaves like the 10-hour half-life row in the calculator.
- Smoking. Tobacco smoke induces CYP1A2 and roughly doubles caffeine clearance. When smokers quit, the same coffee suddenly hits harder.
- Hormones. Oral contraceptives about double the half-life, and in late pregnancy it can stretch by 8 hours or more.
- Medicines and alcohol. Heavy alcohol use slows clearance, and some medicines compete for the same enzyme. A pharmacist can check yours.
- Body size. The same 900 mg is about 13 mg per kg for a 70 kg man and about 18 mg per kg for a 50 kg woman.

How would you cut down from 10 cups without the headache?
Cut by about one cup every few days rather than all at once, because stopping abruptly brings withdrawal: headache, poor concentration, fatigue, nausea and low mood that typically start 12 to 24 hours after the last cup and can last 3 to 7 days, by Temple's review.
- Count honestly. Add every tumbler, tea, cola and chocolate for three days.
- Move the last cup earlier. A cut-off around eight hours before bed removes most of the bedtime load, even before you drink less.
- Thin the decoction. Use less decoction and more milk, or a blend with more chicory. The ritual stays; the dose drops.
- Drop one cup a week. Going from ten to five tumblers over a month avoids the withdrawal crash.
- See a doctor if you notice palpitations, chest pain, high blood pressure readings or panic symptoms, or if you are pregnant or on regular medicines.
What's new
- AHA scientific statement: coffee may protect the heart, energy drinks may harm it; up to 400 mg a day judged safe ScienceDaily (American Heart Association), Jul 2026
- FDA adverse-event reports link energy drinks and stimulant products with cardiac arrest in young adults American Heart Association Newsroom, Nov 2026
- FSSAI requires coffee and chicory percentages on the front of coffee-chicory packs from 1 July 2026 GST Times, Feb 2026
What if next?
- What if you ate the world's hottest chilli? Caffeine speeds your heart; capsaicin can briefly clamp your brain's arteries.
- What if you ate no salt for a month? Your kidneys can save sodium better than you think, until you start sweating.
- How big can a dosa get before it collapses? The breakfast that goes with the kaapi, pushed to 37.5 metres and beyond.
For more on Bengaluru's coffee, chai and café culture, see the Rasa world page.
Questions people ask
How many cups of filter coffee a day is safe?
For most healthy adults, about four tumblers a day keeps you under 400 mg of caffeine, the level EFSA, the FDA and the AHA accept. If your decoction is strong or your blend is all robusta, three may already reach it. Pregnancy limits are lower, about 200 mg a day.
Can 10 cups of coffee kill you?
Almost never, if it is brewed coffee spread over a day. Fatal caffeine poisoning usually needs blood levels of 80 to 100 mg per litre, which means swallowing many grams quickly, typically as tablets or powder. Ten tumblers is about 900 mg over 13 hours.
Does chicory reduce the caffeine in filter coffee?
Yes, in proportion to how much of the blend it replaces, because chicory root has no caffeine. A pack must now state the split, and FSSAI sets 51% coffee as the minimum for a coffee-chicory mixture.
How long does one cup of filter coffee stay in your system?
Half of it is gone in about 3 to 7 hours for most adults, and about 90% is gone after three to four half-lives, so roughly 10 to 25 hours. Smokers clear it faster; people in late pregnancy or on the pill clear it more slowly.
Sources
- Temple JL et al.: The Safety of Ingested Caffeine: A Comprehensive Review, Frontiers in Psychiatry (2017) (half-life, CYP1A2, smoking, pill and pregnancy effects)
- Yamamoto T et al.: Autopsy report for a caffeine intoxication case and review of the current literature, Journal of Toxicologic Pathology (2015) (lethal dose and blood levels, mechanism)
- EFSA NDA Panel: Scientific Opinion on the safety of caffeine, EFSA Journal 13(5):4102 (2015) (200 mg and 400 mg limits)
- EUFIC: EFSA opinion on the safety of caffeine (2015) (caffeine per serving of coffee, espresso, tea, cola)
- US FDA: Spilling the Beans, how much caffeine is too much? (2024) (400 mg, 1,200 mg seizure warning)
- ScienceDaily / American Heart Association: Caffeine and Cardiovascular Disease, a scientific statement (Circulation) (2026) (9.4–20.6 mg per fl oz, heart findings)
- American Heart Association: Energy drinks, stimulant products linked to cardiac arrest, heart attack in young adults (2026)
- GST Times: FSSAI mandates clear labelling for coffee-chicory mixes from July 1, 2026 (2026) (51% coffee minimum)
- Honoré de Balzac: The Pleasures and Pains of Coffee (from Traité des excitants modernes), trans. Robert Onopa, Michigan Quarterly Review 35(2): 273–277 (1996) (Balzac's escalating methods and the effects he reported; the story)

