How to Treat Psychosis Without Medication

Psychosis can be treated without medication, but the honest answer is more layered than a simple yes or no. A growing body of research supports several non-drug approaches that reduce psychotic symptoms, and a few alternative care models have achieved outcomes comparable to standard medication-based treatment with far less drug use. That said, most of the strongest evidence positions these approaches as complements to medication rather than outright replacements, and abruptly stopping antipsychotics carries real risks. What follows is a practical look at which non-medication strategies have genuine clinical evidence behind them, what they can realistically accomplish, and how to think about reducing drug reliance safely.

Cognitive Behavioral Therapy for Psychosis

Cognitive behavioral therapy adapted for psychosis, often called CBTp, is the most widely studied psychological treatment for psychotic symptoms. It works by helping people examine and reframe the beliefs driving distressing experiences like paranoia or voice-hearing. Rather than trying to eliminate hallucinations outright, CBTp helps you change your relationship with them: questioning whether a threatening voice is actually dangerous, testing out whether paranoid thoughts match reality, and building coping strategies that reduce distress even if the experiences continue.

The evidence is strongest for people in the early stages of psychosis or those considered at high risk of developing it. A meta-analysis of preventive interventions in people at ultra-high risk found that CBT cut the rate of transitioning to full psychosis by roughly half at twelve months, and the protective effect persisted out to four years of follow-up.¹ When the same analysis looked at other intervention types, only psychological approaches reached that threshold of effectiveness. In people who already have a first episode of schizophrenia, brief CBTp targeting positive symptoms like hallucinations and delusions has also shown improvements.²

CBTp is now recommended in clinical guidelines across several countries as a standard component of psychosis treatment. In the UK, for example, the National Institute for Health and Care Excellence recommends it for everyone experiencing psychosis, not just as an add-on but as a core part of care. The practical barrier is access: trained CBTp therapists remain scarce in many regions, and waiting lists can be long.

Acceptance and Commitment Therapy

Acceptance and commitment therapy, or ACT, takes a different philosophical angle than CBT. Instead of challenging delusional beliefs or trying to reduce hallucination frequency, ACT teaches you to notice psychotic experiences without fighting them and to redirect your energy toward things you value. The idea is that much of the suffering in psychosis comes not from the voices or beliefs themselves but from your struggle against them.

Over the past two decades, ACT adapted for psychosis has shown meaningful effects on rehospitalization rates, depression, and the distress caused by psychotic symptoms.³ A controlled trial focused specifically on auditory hallucinations found significant improvements in hallucination severity both immediately after treatment and three months later.⁴ The evidence is not as deep as for CBTp, but it fills an important gap for people who find the confrontational style of cognitive restructuring unhelpful or exhausting. Some people do better when the therapeutic message is “you can live meaningfully alongside these experiences” rather than “let’s prove these experiences wrong.”

AVATAR Therapy for Persistent Voices

One of the more striking innovations in psychosis treatment involves creating a digital avatar that represents the voice a person hears. In AVATAR therapy, a therapist works with you to design a computer-generated face and voice that matches your hallucinated voice as closely as possible. Then, across a series of sessions, the therapist speaks through the avatar while gradually shifting its tone from hostile and commanding to supportive and conciliatory. The idea is to give you the experience of standing up to your voice and watching it back down.

A randomized trial published in The Lancet Psychiatry found that AVATAR therapy produced significantly greater reductions in hallucination severity compared to supportive counseling after twelve weeks, with a large effect size.⁵ A later meta-analysis pooling six studies confirmed the finding, reporting a moderate reduction in overall hallucination severity and a notable decrease in voice-related distress specifically.⁶ What makes AVATAR therapy particularly interesting is that several of these studies enrolled people whose hallucinations had not responded to medication, including clozapine. If your voices have persisted despite drug treatment, this is one of the few approaches with evidence specifically in that treatment-resistant group.

EMDR and Trauma-Focused Work

There has been a long-standing clinical reluctance to do trauma-focused therapy with people experiencing psychosis, based on the worry that processing traumatic memories could destabilize them further. The evidence increasingly suggests this caution was overblown. Eye movement desensitization and reprocessing, or EMDR, has been tested in several randomized trials with psychosis patients, and no serious adverse events related to the therapy were reported in any of them.⁷

EMDR in psychosis patients was most effective at reducing negative symptoms (things like emotional withdrawal and lack of motivation) and paranoid thinking. It also improved PTSD symptoms, which matters because trauma history is extremely common in people with psychosis. Results for hallucinations and delusions specifically were more mixed, with some trials finding improvements and others not.⁸ Still, the emerging picture is that trauma-focused therapies like EMDR can safely be used in psychosis and may help with both trauma-related and psychotic symptoms, either by targeting traumatic memories directly or by reducing the distress that feeds into psychotic experiences.⁹

Open Dialogue and Soteria Models

Most of the approaches above are individual therapies that slot into conventional treatment. Open Dialogue and the Soteria model represent something more radical: entirely different systems of care where medication is minimized from the start.

Open Dialogue, developed in Western Lapland, Finland, treats first-episode psychosis through immediate, intensive meetings involving the person, their family, and a treatment team. The meetings happen within 24 hours of contact, continue as often as daily, and focus on generating open conversation about the crisis rather than rushing to diagnose and prescribe. Antipsychotics are available but are deliberately not used as a first-line intervention. A nineteen-year follow-up of the program found that people treated with Open Dialogue had significantly lower rates of hospitalization, disability, and antipsychotic use over the long term compared to people treated conventionally, with no difference in suicide rates between the groups.¹⁰

The Soteria model, originally created in the 1970s in California, takes a different approach to the same goal. It uses small, home-like residences where people in acute psychosis live alongside trained non-professional staff. The environment is designed to be calm and non-coercive, with medication used only when someone requests it or when other approaches have clearly failed. A systematic review found that Soteria achieved results equal to conventional hospital treatment across most outcomes, and in some areas better results, with considerably less medication.¹¹ A more recent two-year naturalistic comparison found that people in a Soteria setting showed greater improvement in personal recovery scores, though the advantage became less clear after adjusting for baseline differences in symptom severity.¹²

Both models are genuinely compelling, but they come with a significant caveat: they work as entire systems, not as isolated techniques. You cannot do Open Dialogue in a clinic that is otherwise organized around 15-minute med checks. These approaches require restructured services, different staffing models, and a cultural willingness to tolerate the uncertainty of not immediately prescribing. They remain rare outside of Scandinavia and a handful of pilot sites elsewhere.

Family Interventions and Expressed Emotion

The family environment has a measurable influence on psychosis relapse. Research going back decades has shown that when family members express high levels of criticism, hostility, or emotional over-involvement toward a person with psychosis, relapse rates climb steeply. In a landmark controlled study, the relapse rate among people in high-criticism households who received standard drug treatment alone was about 41 percent in the first year. Adding family psychoeducation dropped it to 19 percent, adding social skills training dropped it to 20 percent, and combining both brought the relapse rate to zero. In households that shifted from high to low expressed emotion during the study, no one relapsed.¹³

Family psychoeducation teaches relatives what psychosis is, how to communicate without escalating conflict, and how to set reasonable expectations. It is not about blame; it recognizes that living with someone in psychosis is stressful and that families need practical tools. The consistent finding across the literature is that reducing expressed emotion in the household lowers relapse risk, sometimes dramatically.¹⁴ If you are supporting a family member with psychosis, this is one of the most evidence-backed things you can do: learn about the condition, moderate your emotional tone, and participate in structured family work if it is available.

Brain Stimulation

Repetitive transcranial magnetic stimulation, or rTMS, uses targeted magnetic pulses to change activity in specific brain regions. For psychosis, the most common approach involves directing low-frequency stimulation at the left temporoparietal cortex, a region involved in auditory processing that tends to be overactive during hallucinations. The goal is to dial down that overactivity.

A randomized trial comparing active rTMS to sham treatment found that active stimulation produced a large reduction in auditory hallucination scores after two weeks, and the improvement held at six weeks.¹⁵ Another study using twice-daily low-frequency rTMS in people with treatment-resistant hallucinations found a roughly 39 percent reduction in hallucination severity on average, including in patients whose hallucinations had not responded to clozapine.¹⁶ Neurofeedback, a related technique where people learn to regulate their own brainwave patterns using real-time EEG displays, has also shown preliminary improvements in negative symptoms of schizophrenia, though the evidence base is still limited to case studies and a small number of controlled trials.¹⁷

Brain stimulation is not a stand-alone replacement for medication in most cases, but it offers a genuinely different mechanism and can help when voices persist despite adequate drug trials. The practical limitation is access: rTMS requires specialized equipment and multiple sessions, and it is not widely available outside of academic medical centers.

Exercise, Sleep, and Lifestyle Factors

Physical exercise does not eliminate hallucinations or delusions, but it targets some of the most disabling and medication-resistant aspects of psychosis: cognitive problems, negative symptoms like apathy and social withdrawal, and metabolic side effects of antipsychotics. A randomized trial found that aerobic exercise improved cognitive function and reduced negative symptom scores in people with schizophrenia.¹⁸ Even in people taking clozapine, which is typically reserved for treatment-resistant cases and carries substantial metabolic risks, structured exercise programs improved cognitive performance and psychological quality of life compared to a control group.¹⁹

Sleep deserves separate attention because it is both a symptom and a driver of psychotic experiences. Research has established that insomnia is not merely a byproduct of psychosis but a contributing cause: treating sleep problems directly can reduce paranoia and hallucinations.²⁰ A feasibility trial of a targeted sleep intervention in young people at ultra-high risk of psychosis found that sleep disruption contributes to the occurrence of psychotic experiences, anxiety, and depression, suggesting that sleep-focused treatment could have downstream benefits on psychotic symptoms themselves.²¹ If you are managing psychosis and your sleep is poor, that is worth addressing directly, through behavioral sleep strategies, not just through sedating medications.

The Ketogenic Diet and Metabolic Approaches

This is a newer area of research that has generated genuine excitement and legitimate caution in roughly equal measure. The hypothesis is that psychotic disorders involve disrupted brain energy metabolism, and that a ketogenic diet, which shifts the brain’s fuel source from glucose to ketone bodies, could help correct this. Mechanistically, ketone bodies appear to improve mitochondrial function, influence several neurotransmitter systems, and reduce inflammation.²²

The first randomized controlled trial of a ketogenic diet in people with schizophrenia-spectrum and bipolar disorders found that after four months, participants on the diet showed improvements in positive symptoms, negative symptoms, depression, and cognitive performance. Blood ketone levels correlated directly with improvement in depressive symptoms, suggesting that ketosis itself, rather than weight loss, was driving the psychiatric benefits.²³ An earlier pilot study had found a roughly 32 percent reduction in psychiatric symptom severity scores among participants with schizophrenia over four months.²⁴

Gut health is also attracting research attention. The gut microbiome differs significantly between people with psychosis and healthy controls, and these differences are associated with symptom severity.²⁵ A double-blind randomized trial of a probiotic formulation in people with bipolar or schizophrenia-spectrum disorders found that probiotics reduced markers of intestinal permeability compared to placebo.²⁶ Whether improving gut barrier function translates into meaningful symptom reduction remains an open question, but the research direction is plausible and being actively pursued.

The key limitation with all metabolic approaches is that the studies remain small and early. The ketogenic diet in particular is extremely difficult to sustain, especially for people already dealing with the cognitive and motivational challenges of psychosis. These are not ready-made alternatives to medication so much as promising adjuncts that may eventually earn a more central role.

Safely Reducing Medication

For many people reading this, the real question is not whether to start from zero but whether to taper down from a current prescription. This is where things get risky if handled poorly. Stopping antipsychotics abruptly can trigger withdrawal symptoms, including insomnia and rebound psychosis, that are sometimes mistaken for a relapse of the underlying illness. Recent research has found that slower tapering, over months or longer, is associated with a lower relapse rate than quick reductions over weeks.²⁷

The most promising tapering strategy is called hyperbolic tapering, where each dose reduction is smaller than the last. Early cuts might reduce your dose by a noticeable amount, but as you get to lower doses, the reductions become tiny. This approach accounts for the fact that at lower doses, each milligram removed has a proportionally bigger effect on brain receptor activity than at higher doses. The strategy has been modeled even for long-acting injectable antipsychotics, where hyperbolic tapering can be achieved by gradually extending the time between injections rather than cutting doses.²⁸ A practical challenge remains: many psychiatric medications are not manufactured in the tiny doses needed for the final steps of a hyperbolic taper, which means compounding pharmacies or liquid formulations may be necessary.²⁹

Anyone considering reducing antipsychotics should do so in close collaboration with a prescriber who understands gradual tapering principles and is willing to go slowly. The non-medication strategies described throughout this article, particularly CBTp, family psychoeducation, exercise, and sleep optimization, become even more important during and after a taper. They provide alternative sources of stability as the pharmacological support decreases. Reduction is not the same as elimination, and for many people a lower dose combined with robust psychosocial support may be a more realistic and safer goal than complete discontinuation.

When Non-Medication Approaches Are Not Enough

Not everyone with psychosis is a good candidate for medication-free treatment. People in acute crisis with risk of harm to themselves or others generally need the rapid stabilization that antipsychotics can provide. Those with a long history of severe relapse after discontinuation may find that the risks of going without medication outweigh the benefits. And some psychotic conditions, particularly those involving severe thought disorganization or persistent command hallucinations, may simply not respond adequately to psychological or lifestyle interventions alone.

The research on non-medication approaches is also weighted toward certain populations: first-episode psychosis, early-stage or ultra-high-risk individuals, and people who are already somewhat stable on medication. Whether these interventions work as well for people with chronic, treatment-resistant illness is less clear, though the rTMS and AVATAR therapy findings in medication-resistant groups are encouraging. If you are considering any of these approaches, the evidence supports them most strongly as part of a comprehensive treatment plan rather than as isolated alternatives, and the decision should be made with a clinician who knows your specific history and risk profile.