Switching antipsychotics

Why the switch method matters

Antipsychotic switches carry three risks that dose adjustments alone don't. First, symptom breakthrough (psychosis returns during the transition). Second, rebound and discontinuation effects: cholinergic rebound (clozapine, olanzapine, quetiapine to a lesser extent), rebound psychosis with rapid discontinuation of high-D2-blocking drugs, and withdrawal dyskinesia. Third, receptor-based side effects during overlap: EPS if you cross-taper from a low-D2-affinity drug (clozapine, quetiapine) to a high-D2-affinity drug (haloperidol, risperidone) without adjusting.

Pick a method that hedges against the risks most relevant to that patient's history.

The three strategies

Cross-taper

Most common. Reduce the outgoing drug in scheduled steps while adding the incoming drug at low dose, titrating up as you go down. Total transition time 2 to 4 weeks for typical switches, 4 to 6 weeks for clozapine, longer for LAIs.

Example: olanzapine 20 mg to aripiprazole 20 mg over 4 weeks.

• Week 1: olanzapine 20 + aripiprazole 5.
• Week 2: olanzapine 15 + aripiprazole 10.
• Week 3: olanzapine 10 + aripiprazole 15.
• Week 4: olanzapine 5 + aripiprazole 20.
• Week 5: aripiprazole 20.

For aripiprazole specifically, some clinicians prefer a plateau (see below) because the partial agonism can produce activation and akathisia if you push the dose too fast.

Plateau (add-on switch)

Start the new drug and titrate it to full target dose while keeping the old drug at its current dose. Once the new drug is at target and has had time to reach steady state, taper the old drug down. Slower than cross-taper but preserves symptom control the whole way through. Preferred for:

• Stable maintenance patients where any breakthrough risks relapse.
• Switching to aripiprazole from a D2-blocker (mitigates rebound activation).
• Switching to clozapine (evidence supports plateau).
• Switching to any LAI (need to wait for LAI to reach steady state).

Example: risperidone 4 mg to aripiprazole 20 mg by plateau.

• Weeks 1 to 2: risperidone 4 + aripiprazole titrated from 5 to 20.
• Weeks 3 to 4: risperidone 4 + aripiprazole 20.
• Weeks 5 to 8: taper risperidone by 1 mg per week to zero.

The trade-off is a period of dual antipsychotic burden. Usually acceptable for a few weeks.

Abrupt switch

Stop the outgoing drug fully, start the new drug the next day at a standard starting dose (with faster titration than usual). Appropriate for:

• NMS or severe hyperthermic reaction.
• Suspected agranulocytosis (clozapine).
• QTc-driven emergency (thioridazine, ziprasidone in a patient with new prolonged QT).
• Severe intolerance where continuation is unsafe.

Downsides: higher risk of rebound psychosis, cholinergic rebound (if coming off clozapine, olanzapine, or high-dose quetiapine), and relapse. Don't use abrupt switching as the default just because it's operationally simpler.

The switch matrix

Special cases

Clozapine

Never stop clozapine abruptly outside a genuine crisis (agranulocytosis, myocarditis, severe toxicity). Rapid discontinuation carries two big risks: cholinergic rebound (nausea, sweating, diarrhea, agitation, insomnia) and rapid psychotic relapse that can be more severe than baseline. When you must taper, do it over 4 to 8 weeks minimum. When adding a new antipsychotic, use plateau: keep clozapine at current dose while titrating the new drug to target, then taper clozapine slowly.

If discontinuation is urgent for medical reasons (myocarditis, ileus, severe metabolic issue) and the patient must come off clozapine within days:

• Consider anticholinergic cover (benztropine 1 to 2 mg BID) to blunt cholinergic rebound, especially GI and cardiac symptoms.
• Start a bridging antipsychotic immediately (olanzapine is often the closest receptor match).
• Monitor closely for relapse.

LAI to oral, or vice versa

Oral to LAI: the LAI takes weeks to reach steady state (varies by product). During that ramp-up, keep the oral on board. Follow the manufacturer's protocol (paliperidone monthly has its own loading regimen; aripiprazole Maintena requires 14 days of oral overlap; risperidone Consta requires 3 weeks of oral overlap because Consta doesn't release meaningful drug until week 3).

LAI to oral: wait for the LAI plasma level to fall enough that the oral drug won't stack. For monthly LAIs, meaningful drug persists for 1 to 3 months after last dose. For paliperidone 6-monthly, exposure persists for many months. Practical approach: start the oral at a modest dose, don't try to add up doses arithmetically, monitor for both breakthrough and side effects.

Olanzapine and clozapine cross-taper

These two share receptor profiles more than any other pair. Cross-tapering between them tends to work smoothly. The metabolic burden accumulates during overlap (both are heavy on weight, glucose, lipids). Keep the overlap as short as clinically safe.

High-potency to low-potency (and vice versa)

Haloperidol to quetiapine: watch for cholinergic side effects that were being masked by haloperidol's minimal anticholinergic profile, plus a shift in EPS pattern.

Quetiapine to haloperidol: expect EPS to appear as D2 blockade increases. Consider prophylactic benztropine 1 to 2 mg BID during the switch if the patient is at high EPS risk.

Withdrawal dyskinesia vs tardive dyskinesia

When antipsychotic dose comes down (during taper, cross-taper, or discontinuation), some patients develop new movements: orofacial dyskinesias, tongue movements, choreoathetoid limb movements. This is withdrawal dyskinesia. It's caused by the unmasking of upregulated D2 receptors as blockade decreases. It usually resolves over weeks to months as the receptors downregulate.

Withdrawal dyskinesia is not tardive dyskinesia. TD is persistent and often permanent. But an antipsychotic taper can also unmask underlying TD that was suppressed by ongoing D2 blockade, so if the movements persist beyond 4 to 6 months after the taper is complete, that's TD until proven otherwise.

Practical approach when new movements appear during a taper:

• Document with AIMS.
• Slow the taper (don't necessarily abandon it).
• If movements are severe, consider re-uptitrating temporarily and then tapering more slowly.
• Consider VMAT2 inhibitors (valbenazine, deutetrabenazine) if movements persist and cause functional impairment.

Rebound psychosis

Rapid discontinuation of high-D2-affinity antipsychotics (haloperidol, risperidone, olanzapine, paliperidone) can produce psychotic symptoms that are more severe than the patient's baseline. This is thought to reflect D2 receptor upregulation during chronic blockade. It's most reliably documented with typicals but occurs with SGAs too.

The clinical implication: don't abruptly stop a high-D2 blocker unless you have to. Cross-taper or plateau. If you must stop abruptly, expect a rougher few weeks and consider bridging with a different antipsychotic (or benzodiazepines short-term for agitation).

Supersensitivity psychosis is a related but more contested concept: chronic dopamine blockade produces receptor upregulation that manifests as psychosis that's harder to treat than the original illness. It's more commonly discussed with long-term typicals. The evidence is mixed, and the clinical implication is the same either way: taper slowly.

Metabolic switching

The evidence for improving metabolic profile by switching antipsychotics is best for aripiprazole. Switching from olanzapine (or clozapine, or higher-metabolic-risk drugs) to aripiprazole produces meaningful improvements in weight, glucose, lipids, and prolactin. The trade-offs: activation, akathisia, and (less commonly) breakthrough of underlying symptoms.

The Correll group and others have documented that most metabolic gains from an olanzapine-to-aripiprazole switch happen in the first 3 to 6 months. It's not a solution for everyone: patients whose symptoms are stably controlled on olanzapine and who don't tolerate aripiprazole may be better served by staying on olanzapine plus adding metformin (Carolan et al 2025 supports metformin as add-on) and other metabolic interventions.

Lurasidone and ziprasidone are the other reasonable metabolic switches. Cariprazine and brexpiprazole (both partial agonists) are also metabolic-friendly but the switch evidence is thinner.

Common questions

How fast can I switch from olanzapine to aripiprazole? Cross-taper over 4 weeks is standard. Plateau (add aripiprazole to target while keeping olanzapine at current dose, then taper olanzapine) is smoother if the patient is stable and can tolerate a few weeks of dual therapy. Expect activation and possible akathisia in the first 2 weeks of aripiprazole; a slower aripiprazole titration (5, 10, 15, 20 over 2 weeks) helps. Don't stop olanzapine before aripiprazole is at target: aripiprazole's partial agonism can feel undermedicated during the ramp.

Do I need to overlap when switching between LAIs? Usually yes, but the form depends on the LAIs involved. Paliperidone monthly to Trinza (3-monthly): direct switch per manufacturer protocol at the time of the next scheduled dose. Aripiprazole Maintena to Aristada: switch at next scheduled dose using the equivalent Aristada strength. Risperidone Consta to Invega Sustenna: give the Sustenna at the time the Consta would have been due (Consta's release profile means the last Consta dose is still active for weeks). Cross-molecule LAI switches often need bridging with oral for a few weeks.

When is abrupt switch appropriate? Emergencies. NMS, agranulocytosis, myocarditis, severe cardiac events (QT emergency, arrhythmia), severe hyperthermia, severe dystonia unresponsive to treatment, life-threatening allergic reactions. Also when the current drug is being stopped for reasons of confirmed harm (positive tox on clozapine adherence testing that raises safety concerns, etc.) and you can't tolerate any continued exposure. Everything else: cross-taper.

What's withdrawal dyskinesia and does it mean I've caused TD? Withdrawal dyskinesia is new abnormal movements that appear during antipsychotic dose reduction and typically resolve over weeks to months. It reflects unmasking of D2 receptor upregulation. It's not the same as tardive dyskinesia, though the mechanism overlaps. If movements persist beyond 4 to 6 months after the taper is complete, that's TD by consensus definition. During a taper, if you see new movements, document with AIMS, consider slowing the taper, and re-evaluate. A brief return of movements during dose reduction doesn't retrospectively mean you caused persistent TD.

Should I stop clozapine before starting a new antipsychotic? No, unless discontinuation is urgent for safety. Plateau is the preferred approach: add the new antipsychotic at target dose, let it take effect, then taper clozapine slowly (4 to 8 weeks). Rapid clozapine discontinuation risks cholinergic rebound and rapid, sometimes severe, psychotic relapse. Even in urgent scenarios, consider anticholinergic cover to blunt the cholinergic rebound.

Sources

• Maudsley Prescribing Guidelines in Psychiatry, 14th edition.
• American Psychiatric Association. Practice Guideline for the Treatment of Patients with Schizophrenia, 3rd edition (2020).
• Kishimoto T, Hagi K, Nitta M, et al. Effectiveness of long-acting injectable vs oral antipsychotics in patients with schizophrenia: meta-analysis. Schizophr Bull. 2018;44(3):603-619.
• Correll CU, Kishimoto T, Kane JM. Metabolic side effects of second-generation antipsychotics and their management. World Psychiatry.
• Carolan A, et al. Metformin for antipsychotic-associated weight gain: pragmatic update. Schizophr Bull. 2025.
• Chouinard G, Chouinard VA. New classification of selective serotonin reuptake inhibitor withdrawal. Psychother Psychosom. On supersensitivity psychosis concept.
• Package inserts for clozapine, aripiprazole Maintena, Aristada, Invega Sustenna, Trinza, Hafyera, risperidone Consta, Uzedy, Perseris.

Reviewed against current guidelines as of June 8, 2026. This is not medical advice.