What Is Hyperbolic Tapering? Why the Last Few Milligrams Are the Hardest

Here is a pattern that repeats constantly in antidepressant tapering. Someone comes down from 20mg to 15mg without much trouble. Then 15mg to 10mg — fine. Then 10mg to 5mg, still manageable. Then 5mg to zero, and it falls apart.

They assume they did something wrong at the end. They didn't. The schedule was wrong from the beginning, and it only revealed itself at the bottom.

Hyperbolic tapering means reducing the dose by a percentage of your current dose rather than by a fixed number of milligrams, so the size of each cut shrinks as the dose falls. It exists because the relationship between antidepressant dose and its effect on the brain is not a straight line — the effect rises steeply at low doses and flattens at higher ones. A fixed 5mg cut therefore removes far more drug effect at the bottom of the range than at the top.[1]

The curve that explains everything

Antidepressants like SSRIs work in large part by occupying the serotonin transporter. The critical fact is how occupancy responds to dose: it climbs very steeply from zero, then flattens out.

Practically, this means the difference between 0mg and 5mg represents a large change in the drug's actual effect on your brain, while the difference between 15mg and 20mg represents a comparatively small one — even though both are 5mg on paper.

This relationship was set out by Horowitz and Taylor in Lancet Psychiatry in 2019, and it is the pharmacological basis for the taper schedules in the Maudsley Deprescribing Guidelines.[1]

What this looks like in practice

Compare two schedules starting from 20mg.

Linear (the intuitive one): 20 → 15 → 10 → 5 → 0. Four equal-looking 5mg steps. In terms of drug effect, the final step is by far the largest — which is exactly why people crash at the end.

Hyperbolic (proportional): each step removes about 10% of the current dose. 20 → 18 → 16.2 → 14.6 → 13.1, continuing down through 5mg, 2mg, 1mg and below. The milligram cuts get smaller and smaller. Each step feels roughly the same size to your nervous system the whole way down.

The second schedule has many more steps and takes considerably longer. That is not inefficiency — it is the point. You can see what it looks like for your own dose in the taper schedule builder.

The part nobody warns you about: your tablets run out

Proportional tapering runs into a practical wall. Reducing by 10% of 2mg means a 0.2mg step, and no tablet splits that finely.

This is where the formulation of your specific medication becomes the limiting factor:

• Manufactured oral liquids exist in the US for sertraline (20mg/mL), escitalopram (1mg/mL), citalopram (2mg/mL), fluoxetine (20mg/5mL) and paroxetine (10mg/5mL). Where one exists, this problem is largely solved — ask your prescriber to write for it.
• Venlafaxine XR and duloxetine have no US liquid, and their capsules cannot be split or crushed. Options there are switching to a splittable form, counting capsule beads under prescriber supervision, or compounded doses.
• Bupropion XL and mirtazapine have their own constraints worth discussing individually.

A good schedule tells you in advance where this wall is, so you can arrange a solution before you hit it rather than discovering it at the pharmacy counter. The builder flags the exact step where your doses drop below what standard forms can deliver.

How slow is slow?

Slower than most package inserts imply, and slower than most people expect.

Reaching a very low final dose with 10% steps takes roughly 38 reductions. At a two-week hold that is around 18 months; at four weeks, closer to three years. Those numbers genuinely surprise people.

Two things make them less daunting:

1. Most people do not run the whole schedule at one speed. Larger steps early and small steps at the low end is a common and reasonable pattern to agree with your prescriber — the curve is flat up top, so bigger cuts there cost you less.
2. What matters is that each individual step is small enough to be uneventful. A slow taper you complete beats a fast one you abandon and restart.

The habit that matters most

Judge every reduction at the end of its hold, not in the first few days.

Antidepressant levels take days to re-stabilise after a change, and longer for drugs like fluoxetine. A step that feels fine on day two can catch up with you on day six; a step that feels rough on day three has often settled by week three. Deciding on day three means reacting to the worst point of a curve you haven't seen the end of.

If a step brings symptoms that do not settle, hold at that dose rather than pushing on. Holding is a normal part of a successful taper, not a setback. Tracking day to day makes this much easier to judge — the symptom tracker is built for exactly that.

A note on what this does not tell you

Hyperbolic tapering is a method for how to reduce. It says nothing about whether you should.

That decision belongs to you and your prescriber, and it depends on things no schedule can know: your history, how many episodes you have had, what else is going on in your life right now, and how well the medication is working. Depression or anxiety returning during a taper is a real possibility and needs a clinical response, not a smaller dose step.

Nothing here is medical advice. Every schedule this site produces is a proposal to discuss with your prescriber — never a substitute for one.