Ashton Method vs. Hyperbolic Tapering: Where the Two Approaches Disagree

For over twenty years, anyone serious about coming off a benzodiazepine eventually found the Ashton Manual — Professor Heather Ashton's 2002 guide built from her Newcastle withdrawal clinic. It remains the document that taught the world that slow beats fast.

But it is no longer the only serious playbook. In June 2026, a group led by Dr. Mark Horowitz — with co-authors including patient advocates and the authors of the Maudsley Deprescribing Guidelines — published the first peer-reviewed, benzodiazepine-specific case for hyperbolic tapering in Psychological Medicine. And a year earlier, the American Society of Addiction Medicine released its joint clinical guideline on benzodiazepine tapering with nine partner societies.

The short version: everyone now agrees on "go slow, never stop abruptly, let symptoms set the pace." The disagreements are about the shape of the taper and whether you should switch to diazepam first. Those two disagreements are worth understanding before you and your prescriber pick a plan.

What the Ashton Method says

The classic Ashton protocol has two pillars:

1. Crossover to diazepam. Because diazepam has the longest half-life of any benzodiazepine (up to ~200 hours with its active metabolite) and comes in small tablet strengths and a liquid, Ashton moved most patients from their shorter-acting drug (alprazolam, lorazepam, clonazepam) onto an equivalent dose of diazepam in stages.
2. Stepwise reductions on a published schedule. Her manual's schedules reduce in fixed milligram steps — roughly 5–10% of the dose every 1–2 weeks at the start — finishing from around 0.5–1 mg of diazepam, with the whole process typically taking around 10 months or more.

It worked for enormous numbers of people, and its core insights — slow, patient-led, never cold turkey — have now been absorbed into every mainstream guideline.

What hyperbolic tapering says

The Horowitz group's argument starts from receptor biology. The relationship between benzodiazepine dose and GABA-A receptor occupancy is hyperbolic: the first few milligrams produce most of the effect, and each additional milligram adds less. Which means the reverse is also true — at low doses, each milligram you remove takes away a disproportionately large amount of effect.

The practical consequences, per the 2026 paper:

• Cut by a percentage of your current dose, not your original dose. A "10% taper" in the Ashton style means 10% of where you started — the same milligram amount every step, which becomes a progressively bigger biological cut as you descend. Hyperbolic tapering takes 5–10% of whatever you're on now, so the milligram steps shrink continuously and each step removes a roughly constant slice of receptor effect.
• The final doses are much smaller than most schedules assume. The paper argues final doses as low as 0.2 mg diazepam-equivalent or less may be needed before jumping to zero — well below the 0.5–1 mg endpoint of classic schedules, and far below the smallest tablet.
• Crossover to diazepam is optional, not required. Staying on your original drug — using a manufacturer's liquid, a compounded suspension, or careful tablet techniques — avoids the risks of the switch itself (equivalence is approximate, and some people tolerate the crossover badly). Diazepam remains a reasonable tool, especially for very short-acting drugs; it's just no longer the default first move.
• Micro-tapering is legitimate. Daily microgram-scale reductions, liquid titration, and compounded doses — techniques the patient community refined for years on forums — get explicit endorsement as precision tools.

If that logic sounds familiar, it's because our taper calculator has always worked this way: proportional cuts of the current dose, with the steps shrinking as you descend.

Where ASAM lands

The 2025 ASAM joint guideline — summarized for family physicians in American Family Physician — sits somewhere between the two. It recommends starting with 5–10% reductions, generally not exceeding 25% every two weeks, slowing to 5–10% every 6–8 weeks for strongly dependent patients, and it conditionally supports diazepam crossover for suitable patients. It also states plainly that clinicians should not abruptly discontinue a physically dependent patient — the single most important sentence in the document for anyone whose prescriber is in a hurry.

So the current landscape looks like this:

Question	Ashton (2002)	ASAM (2025)	Horowitz et al. (2026)
Overall pace	Slow, ~10+ months	Slow, months to years	Slow, months to years
Cut size	~5–10%, fixed mg steps	5–10%, review each step	5–10% of current dose
Step shape	Linear (same mg each time)	Not specified	Hyperbolic (shrinking mg)
Diazepam crossover	Default for most	Conditional option	Optional, not default
Final dose before stopping	~0.5–1 mg diazepam	Not specified	As low as ~0.2 mg equivalent
Micro-tapering / liquids	Diazepam liquid	Not addressed	Explicitly endorsed

How to think about the disagreement

A few honest observations:

The disagreement is real but narrower than it looks. Ashton's schedules already made cuts roughly proportional at the top of the range, and she told patients to slow down near the end. Hyperbolic tapering formalizes mathematically what good Ashton-style practice did by feel.

The evidence base for all of it is thin. A 2025 BMJ meta-analysis of 49 randomized trials found low-certainty evidence for most deprescribing interventions — and no trial has directly compared linear against hyperbolic benzodiazepine tapering. ASAM's authors themselves note most of their recommendations rest on expert consensus. Nobody should claim their method is proven superior; what's changed is that the receptor-occupancy argument for shrinking steps now has peer-reviewed, benzo-specific backing.

The practical convergence matters more than the doctrinal fight. Every current authority agrees on: no abrupt discontinuation, patient-led pace, small cuts that get smaller near the end, and holding when symptoms flare. If your plan honors those four things, you are inside the standard of care under any of the three frameworks.

Where it changes what you actually do: if you're near the end of a taper and each cut is hitting harder than the last, the hyperbolic framing explains why — and the fix is smaller cuts (which may require liquid or compounded doses), not more willpower. And if a crossover to diazepam has been presented to you as mandatory, it isn't; it's one tool with real trade-offs your prescriber should discuss.

Build your own schedule either way

Our taper schedule calculator generates proportional (hyperbolic-style) schedules for any benzodiazepine and shows exactly where doses fall below the smallest tablet. The conversion calculator handles diazepam-equivalence if you and your prescriber do choose a crossover. And the full research behind this article — the 2026 Horowitz paper, the ASAM guideline, the BMJ meta-analysis, and the protracted-withdrawal literature — lives on our research page, which we keep current as new studies publish.

This article is education, not medical advice. Never change your dose or schedule without your prescriber.