Protocol · Evidence-based · Gradual
Quit caffeine permanently — the step-down "don't buy it" method
Cold-turkey quitting is why most people fail: the withdrawal headache hits ~50% of people who stop and peaks around a day or two in. This protocol sidesteps that by tapering the dose your body sees, using the simplest lever there is — not re-buying it. When it's not in the house, the taper happens on its own.
Not medical advice. If you're pregnant, have a heart condition, or take medication affected by caffeine, taper under a clinician's guidance. Take the self-test first →
Why step down instead of quitting cold
Caffeine withdrawal is a real, well-characterised syndrome — enough so that it's a diagnosis in the DSM-5. A review of 66 studies validated ten symptoms (headache, fatigue, low energy and alertness, drowsiness, low mood, poor concentration, irritability, fog, flu-like feelings). Crucially, incidence and severity rise with the daily dose, and symptoms appear from intakes as low as 100 mg/day. Onset is 12–24 h after your last dose, peaks at 20–51 h, and lasts 2–9 days.1 Lowering the dose your body is used to before you reach zero shrinks that peak.
Avoiding withdrawal is itself one of the main reasons people keep using caffeine day to day1 — so the taper also breaks the loop, not just the chemistry.
The 6-step protocol
-
Measure your baseline
Take the self-test and note your estimated mg/day. That number is your starting point; every step is a cut relative to it. Write down which cups are "anchor" habits (the wake-up one, the after-lunch one).
-
Stop re-buying the strong stuff
Finish what's open, then don't replace it. When you next shop, buy one step weaker (see the taper table) — or simply less of it. What's not in the cupboard can't be over-poured.
-
Cut ~25% of the dose each week
Reduce the caffeine your body sees by about a quarter per week: fewer cups, smaller cups, or blending in decaf. A gentle glide over 4–6 weeks keeps you under the threshold where withdrawal bites.1
-
Protect your sleep first
Move your last caffeine earlier every week. In a controlled trial, 400 mg within 12 h of bed measurably disrupted sleep, while 100 mg up to 4 h before bed did not — so cutting the afternoon/evening dose earliest gives the fastest sleep win.7 Better sleep then makes the rest of the taper easier.
-
Blend down to decaf, then to zero
For the last stretch, half-caf → mostly-decaf → herbal or water keeps the ritual (warm mug, morning pause) while the dose approaches nothing. Keeping the ritual is what stops relapse buying.
-
Stay off — keep it out of the house
Permanence is an environment problem, not a willpower one: if you never re-buy it, the default is caffeine-free. Keep decaf/herbal stocked so the ritual has somewhere to go. Re-take the self-test at 4 and 8 weeks to watch your scores fall.
Example glide path (from ~4 cups/day)
| Week | Target | Approx mg/day | |
|---|---|---|---|
| 0 (baseline) | 4 cups | ~380 | |
| 1 | 3 cups | ~285 | |
| 2 | 2 cups + ½ decaf | ~215 | |
| 3 | 2 small | ~150 | |
| 4 | 1 + half-caf | ~95 | |
| 5 | mostly decaf | ~40 | |
| 6 | decaf / herbal | ~0 |
Illustrative only — match the pace to how you feel. Slower is fine; the goal is staying under your withdrawal threshold.1
What you gain — and the evidence for it
Realistic, sourced expectations. Individual results vary; these are group findings from controlled studies and meta-analyses.
Removing a 200–300 mg caffeine dose lowered systolic/diastolic blood pressure acutely by about 8 mmHg and 6 mmHg in a meta-analysis of hypertensive adults — the acute pressor effect disappears once caffeine is gone.
Mesas 2011, meta-analysis. [9]In a randomized trial, several weeks of coffee abstinence lowered systolic blood pressure by about 6 mmHg versus continuing filtered coffee in young normotensive adults.
Bak & Grobbee 1990, RCT. [8]Caffeine reliably raises cortisol; daily use only partly blunts this. After 5 days without caffeine, the cortisol response returns — evidence your stress-hormone system re-sensitises when you stop, rather than staying chronically nudged upward.
Lovallo 2005, RCT crossover. [5]Caffeine lengthens the time to fall asleep, cuts total sleep and slow-wave (deep) sleep, and lowers sleep quality — dose- and timing-dependent. Removing it restores the sleep architecture it was suppressing.
Clark & Landolt 2016, systematic review; Gardiner 2025, RCT. [6][7]Caffeine is anxiogenic at higher doses and strongly so in sensitive people: ~480 mg induced panic attacks in ~54% of panic-disorder patients versus ~2% of healthy controls, and raised anxiety in healthy adults too. Lowering the dose removes that driver.
Klevebrant & Frick 2021, meta-analysis. [4]Handling the first two weeks
If a headache starts
It means you dropped faster than your threshold liked. Hold this week's dose steady (don't go back up), hydrate, and if needed cut the next step in half. Headache incidence and severity scale with the size of the drop.1
Keep the ritual, lose the drug
Most of coffee's pull is behavioural — the warm mug, the pause, the routine. Move that ritual onto decaf or herbal tea so you're not fighting the habit and the chemistry at once.
Expect a mood dip, then a lift
Low mood and irritability are on the validated withdrawal list1 and are temporary. Cross-sectional data links caffeine-use-disorder and cessation with short-term higher stress/anxiety scores — another reason to glide down rather than lurch.10
Re-test to see progress
Re-take the self-test weekly. Watching your withdrawal, anxiety and dependence percentages fall is the most motivating feedback there is — and you can download each result to compare.
Sources
According to PubMed. Every figure on this page traces to one of these peer-reviewed sources; DOIs link to the original article.