Why does the Ashton Manual want you on diazepam in the first place?
Because diazepam leaves the body slowly, and a slow exit is easier to taper. In Chapter 2 of the Ashton Manual, Professor C. Heather Ashton explains that diazepam has a half-life of up to 200 hours once its active metabolite is counted, while people on shorter-acting drugs such as alprazolam and lorazepam often feel a “mini-withdrawal” between doses, known as interdose withdrawal.
Diazepam also comes in small scored tablets and as a liquid, so tiny reductions are easier to make, and the Manual has the switch made gradually, one dose at a time. Clonazepam is often lumped in with the short-acting drugs, and it does not belong there: Table 1 lists it at 18 to 50 hours, against 6 to 12 for alprazolam, so the case for crossing it to diazepam rests on potency and coarse tablet steps, not fast clearance.
Dr. Leeds prefers the diazepam crossover taper for most patients, and in his experience those who can use it tend to have a more comfortable taper. Yet, a preference for most is not a rule for all, and the Maudsley Deprescribing Guidelines, in their chapter on benzodiazepines and z-drugs, likewise treat the switch as something that can help some people, not a required first step. Why does it fail for some, and what do you do then?
Who tends to run into trouble with the crossover?
When the crossover fails, people tend to assume that they misread the Manual or that something is wrong with them. Neither is true. A protocol that has been photocopied enough times starts to look like a law of nature, and the crossover has been photocopied more than most, so Dr. Leeds treats it as one tool among several.
In his experience the people who struggle fall into a few patterns, and the first is the person already destabilized by tolerance withdrawal, a prior failed taper, or an abrupt stop and restart. The community word for this is kindling, each rough withdrawal seeming to leave the nervous system more sensitive to the next, and a kindled nervous system tends to treat a crossover as one more withdrawal.
The second is the person who is too far along. The crossover is most reliably useful early, before the dose has come down much, and the other case turns up often: a patient many years on clonazepam and already far down after a hard year is handed the crossover because the Manual says so, and three weeks later is sleeping two hours a night.
The third is metabolism. Diazepam carries an active metabolite, desmethyldiazepam, which Table 1 lists at 36 to 200 hours, and Professor Ashton notes in Chapter 1 that long-acting benzodiazepines accumulate with repeated dosing, and that older people accumulate them more readily.
Some people, in Dr. Leeds’ experience, process diazepam so slowly that it sedates them, and a few clear it too quickly for it to smooth anything. The fourth is the simplest: an allergy or an adverse reaction to diazepam or a related benzodiazepine takes the crossover off the table.
A fifth idea circulates and deserves honest labeling. One theory holds that a nervous system shaped by years of clonazepam or alprazolam may not read diazepam as the same signal even at a calculated equivalent dose.
Neither the Ashton Manual nor Maudsley describes such a mechanism, so Dr. Leeds offers it as one theory and nothing more. Professor Ashton did, however, concede the larger point herself.
Chapter 2 notes that some people have particular difficulty switching from clonazepam to diazepam, and her answer was specially made capsules containing very small doses of the original drug, used to make gradual reductions. That is the same-medication taper by another name, and it comes from the author of the protocol.
What does “no relief” actually look like?
Usually one of a few things. The symptoms stirred up by the switch do not settle, or the original symptoms remain at full strength, or new ones appear, or you feel more sedated without feeling more stable, with sleep, energy, and mood all worse than before.
How long should you wait? In fact, the Ashton Manual gives no adjustment period, and any number you see in a forum is a guess.
In Dr. Leeds’ experience the settling, when it comes, arrives over a matter of weeks, and if nothing has improved by then, more time alone is unlikely to fix it. While diazepam’s benefit sometimes shows up later in a taper, that is a different picture from feeling steadily worse.
Some of what gets called “no relief” is a dose problem rather than withdrawal. Heavy sedation, confusion, slurred speech, falls, or slowed breathing after a switch mean the diazepam dose is too high for you, and the prescriber needs to hear about it that day.
A seizure, chest pain, a sudden loss of vision or hearing, or thoughts of harming yourself are never a taper decision. Call 911, or, if the thoughts are of self-harm, call or text 988, the 988 Suicide and Crisis Lifeline.
Unfortunately, the people around you may misread a failed crossover as evidence that the problem is you rather than the plan. Physical dependence is not addiction. Dependence is the expected adaptation of a nervous system to a medication taken as prescribed over time, which the 2020 United States Food and Drug Administration (FDA) benzodiazepine label update recognizes in calling for gradual reduction, and a nervous system that objects to a change of molecule is showing that adaptation, not a character flaw.
Hold first, before anyone changes anything else.
So, the first move when the crossover is not working is usually no move at all. A person who has destabilized mid-crossover generally does better holding at the current combination, whatever it is, until the nervous system has a few steady weeks behind it. Professor Ashton allowed for this when she wrote that you can stand still and take a vacation from further withdrawal.
Holding is not failure. A hold gives the nervous system the predictability it is asking for, and Dr. Leeds treats a hold as protective when symptoms spike. What a hold never means is stopping: no benzodiazepine is stopped abruptly, and the Manual is blunt that abrupt or over-rapid withdrawal can cause convulsions.
What are the options once you have held?
Three, and each is a decision made with your prescriber rather than at the kitchen table. The first is partial substitution, keeping the original benzodiazepine and adding a smaller share of diazepam than a full crossover would use. While the mix is less standardized and needs closer monitoring, it gives some of the smoothing without forcing a complete change of molecule, and it suits some people.
The second is returning to the original drug. If symptoms became clearly worse during the crossover and have not improved through a hold, going back is often the cleanest path, and the dose you return to is the prescriber’s call, because a partial crossover reversed by guesswork can leave you well over or under where you started. The third is a direct taper of the original drug.
Alprazolam is the special case. Its half-life of 6 to 12 hours makes a same-medication taper rough, so when diazepam cannot be used Dr. Leeds’ fallback is a crossover to clonazepam, at 18 to 50 hours, and a taper from there.
Professor Ashton preferred diazepam as the destination drug. Yet, when diazepam is off the table, the choice is between a half-life measured in hours and one measured in a day or two, and the longer one wins.
How is a same-medication taper run well?
On the principles the crossover was built on, starting with stabilization: the dose and schedule are held steady until the nervous system feels predictable, because a shaky start makes every later cut harder. Then comes precision, which usually means a compounded liquid, since tablets cannot deliver the very small reductions a slow taper needs.
The third principle is hyperbolic pacing: the cuts get smaller as the dose gets lower, because the last portion of the dose carries the largest share of the effect, and some patients do best with micro-tapering, very small reductions every few days instead of a larger monthly step.
The fourth is consistency: a switch between brands or generic manufacturers mid-taper can change absorption, a sensitized nervous system notices, and an involuntary switch at the pharmacy is worth reporting to your prescriber.
The fifth is the timeline, which follows the body and not the calendar. Professor Ashton wrote that whether a taper takes 6, 12, or 18 months is of little significance after years of use, and that the best judge of the pace is you. Many tapers run a year or more, a shorter one after brief use is not a failure of nerve, and none of this is a schedule.
The protocol is a starting point, not a verdict.
What unites the tapers that go well is a willingness, on both sides of the desk, to treat what your nervous system reports as information. Of course, when diazepam works, the crossover taper remains the route Dr. Leeds prefers. When it does not, the plan changes, and you are not asked to endure a plan for the sake of the plan.
Fortunately, none of the alternatives is a lesser road. Each rests on the slow, patient-directed principles the Manual laid down, and each has carried people to the end of a taper who could not use the famous route. Your part is to report plainly what your body is doing, week by week, and your prescriber’s part is to keep adjusting the route and the pace until the two of you find the taper your body can do.
This article is educational. It is not medical advice, and reading it does not create a doctor-patient relationship. Decisions about starting, continuing, or tapering any medication should be made with your own physician.