Cold Turkey vs NRT: What the Evidence Says

The most common belief about smoking cessation, repeated to me by clients who have already failed several quit attempts, is that cold turkey is somehow the more honest or more legitimate way to stop. The framing is that medication is a crutch and that a real quit means doing it on willpower alone. The framing is wrong. The published evidence on what works is consistent across decades of randomised trials, large cohort studies and Cochrane reviews, and it is not ambiguous.

This article is the doctor-led summary of what the evidence actually says, what the cultural belief gets wrong, and how to make the decision before Day 1. It is the practical companion to the withdrawal timeline, which describes the curve the dose decision rides on top of.

The numbers, in plain English

The published figures are remarkably stable across studies.

Cold turkey, defined as a quit attempt with no pharmacological support and no structured behavioural programme, has a twelve-month success rate of roughly three to five percent. That is for unaided quit attempts in adult smokers. Most cold-turkey relapses happen in the first two weeks; a long-tail remainder fall away through the first three months.

NRT (nicotine replacement therapy) roughly doubles that rate. The published Cochrane numbers put the single-product NRT effect at around 1.7 to 2.0 times the cold-turkey baseline, depending on the product and how the studies are pooled. The same review notes that combination NRT, meaning a long-acting patch plus a short-acting gum or lozenge for breakthrough cravings, outperforms either alone.

Varenicline (the prescription medication, sold under the trade name Champix or Chantix in many markets) performs similarly to or slightly better than combination NRT. The mechanism is different: it sits on the same receptor as nicotine and dampens the reinforcement signal without delivering nicotine itself.

Pharmacological support combined with structured behavioural coaching roughly triples the cold-turkey rate. This is the highest-success path in the published literature. The medication handles the chemistry; the coaching handles the trigger map, the social architecture, the identity work that the chemistry alone cannot reach.

The translation: of one hundred smokers attempting cold turkey, three to five are quit at the twelve-month mark. Of one hundred using NRT, seven to ten. Of one hundred using combination NRT plus structured behavioural coaching, ten to fifteen. The absolute numbers are sobering across the board. The relative differences are not.

Why the "willpower is more legitimate" framing is wrong

The framing has three sources, and all three are wrong for the same underlying reason.

The first source is the puritan one. The idea that medication for an addiction is morally inferior to overcoming the addiction without medication is a cultural inheritance, not a clinical position. No equivalent framing applies to other chronic-disease treatments. Nobody describes insulin as a crutch for diabetes. The framing is unique to addiction medicine, and it has cost real quitters real outcomes.

The second source is the misunderstanding of what nicotine receptor downregulation requires. The cravings on Day 3 are not a test of character; they are a receptor adaptation. The receptor density rebuilds whether nicotine is delivered through a cigarette, through a patch, or through a lozenge. NRT slows the rebuild because it keeps the receptor partially occupied at a lower steady dose, but the rebuild still happens; the receptors are returning to non-smoker density over the same three weeks. The framing that a patch is "still smoking" is wrong on the chemistry.

The third source is the relapse-shame loop. A quitter who relapses on a cold-turkey attempt often concludes that they lacked the willpower. The conclusion entrenches the framing. The published data tells a different story: the cold-turkey path is the lowest-success path because the receptor adaptation outpaces unaided behavioural control on Day 3 for most adult smokers. The relapse is a chemistry event, not a character event. Renaming it as a character failure has cost most of my clients the willingness to try the dose that would actually work for them.

When does cold turkey work

Cold turkey is the lowest-success path on average, but it is not zero. A subset of quitters do quit clean on Day 1 with no support and stay quit. The published profile of that subset has three rough features. They are typically light or social smokers with a recent and short history. They typically have a major external life event — a pregnancy, a serious health scare, a strong family pressure — that supplies extrinsic motivation across Day 3 and Day 14. And they typically have a high baseline of behavioural self-regulation that the rest of their life already reflects.

If you are in that profile, cold turkey is a reasonable option to try. If you are not, the published rate of three to five percent should be read as what it is: most quit attempts in that subset fail, and most failed attempts then get retried with the dose, which works. The cleaner sequence is to skip the first attempt and use the dose from Day 1.

What "the dose that works" actually looks like

For a typical adult cigarette smoker, the evidence-based starting point is combination NRT: a 21mg patch for moderate-to-heavy smokers (more than ten cigarettes a day) or a 14mg patch for lighter smokers, plus a short-acting gum or lozenge for breakthrough cravings. The patch goes on first thing in the morning. The short-acting form is used when the craving arrives, not on a schedule.

The published full programme is eight to twelve weeks, with a step-down at week four to six and another at week eight to ten. The full programme matters; quitters who stop the patch at week three to "test themselves" relapse at a noticeably higher rate than those who taper through the published schedule.

For vape, shisha, midwakh and dokha users, the dose tables need to be adjusted. The published cigarette tables undersedate non-cigarette nicotine users for reasons covered in the substance-specific articles (vapes, shisha, midwakh, dokha). The starting NRT dose for those users is typically at the upper end of the published range or above, and the dose conversation in the first appointment is the one that prevents undersedation in week one.

Varenicline is the alternative prescription path. Mechanism: the molecule binds the receptor as a partial agonist, reducing both the craving and the reward signal if you smoke through it. The published efficacy is similar to or slightly better than combination NRT. Side effects in the first week include vivid dreams; in a small subset, low-mood symptoms that warrant a clinical conversation. Dose escalation across the first week is standard practice. Discuss with your pharmacist or GP whether varenicline is appropriate for your clinical picture.

The behavioural side

The medication handles the chemistry. It does not handle the cafe, the wedding, the friend group, the post-meal moment, or the grammar you use to describe yourself at Day 30. Those are the surfaces the structured behavioural work covers, and they are most of why the combination of medication and coaching outperforms either alone.

The published evidence on what coaching contributes is consistent. Pre-positioned substitute behaviours for the predictable triggers. The trigger map across the first month. The Cafe Rebuild for the social setting. The grammar of quitting for the identity shift at Day 28. The Day 14 dopamine-dip recognition. Each of these is a small intervention in isolation; the package is what compounds.

What to do this week

Five things, in order:

1. Decide the dose path now, before Day 1. Combination NRT for most adult smokers. Varenicline as the alternative. Cold turkey only if you fit the narrow profile above.
2. Book the pharmacist or GP appointment. The dose conversation is the one that prevents undersedation in week one. Most quitters undershoot this; round the daily cigarette count up if you are uncertain.
3. Pick the quit date. Wednesday or Thursday work best for most adult smokers, with Day 3 falling on a quieter day.
4. Read the timeline. The withdrawal timeline tells you what each day will feel like. Quitters who know the curve in advance handle Day 3 and Day 14 better than those who do not.
5. Plan the behavioural side. The trigger map, the substitute object, the household conversation, the activity protection. The dose without the behaviour is the seven-to-ten path; the dose with the behaviour is the ten-to-fifteen path.

That is most of the work. The rest is the structured coaching version — the trigger map, the NRT calibration, the Cafe Rebuild for the social setting, the Day 21 receptor reset, and the identity work that becomes the load-bearing part by Day 30. That is what the 1:1 programme on this site does.

A note: NRT and varenicline dosing, contraindications, and interactions belong to your pharmacist or GP. The numbers in this article are population averages from the published literature; the curve you actually run depends on your clinical picture. Bring the dose conversation to whoever knows it.