PCP (Phencyclidine): The Neuroscience

Almost no drug has a stranger résumé than phencyclidine. It failed as a surgical anesthetic, escaped onto the street as “angel dust,” and acquired a fearsome reputation for violence — and yet, quietly, it became the single most important chemical tool psychiatry has for understanding psychosis, and the ancestor of ketamine’s rapid antidepressant revolution. The same properties that made PCP a bad medicine made it an extraordinary probe of the brain. This article is education, not medical advice.

The anesthetic that failed — and the field it founded

Phencyclidine — chemically 1-(1-phenylcyclohexyl)piperidine, an arylcyclohexylamine — was synthesized at Parke-Davis in 1956 and looked, briefly, like a breakthrough. As the intravenous anesthetic Sernyl, it produced profound analgesia and a trance-like detachment without the respiratory and cardiovascular depression that made other anesthetics dangerous. The catch appeared on waking. A substantial fraction of patients — figures up to roughly one in five are cited — emerged into delirium, agitation, and hallucination, sometimes for hours (Bey & Patel, Cal J Emerg Med 2007). That “emergence phenomenon” was disqualifying for a surgical drug, and human use was abandoned in the early-to-mid 1960s; a veterinary formulation lingered until the late 1970s.

Rather than discard the chemistry, Parke-Davis mined it. In 1962 the company’s chemists produced ketamine, a shorter-acting, less disturbing homolog that kept the useful anesthesia and shed most of the emergence problem — and which remains in worldwide clinical use today. PCP, meanwhile, escaped into the street as “angel dust.” But its most consequential legacy was neither the operating room nor the drug scene. It was the laboratory: the same properties that made PCP a bad anesthetic made it an extraordinary probe of the brain.

Mechanism: the open channel and the drug that plugs it

The defining discovery came in 1983, when Anis, Berry, Burton, and Lodge showed that ketamine and PCP selectively blocked excitation driven by N-methyl-D-aspartate while sparing other glutamate responses (Anis et al., Br J Pharmacol 1983). This identified the dissociatives as NMDA-receptor antagonists — and specifically as open-channel (non-competitive, “use-dependent”) blockers. The NMDA receptor is a glutamate-gated ion channel; PCP binds a site deep within the open pore (the eponymous “PCP site”), physically plugging it so ions cannot pass. Because the drug can only reach that site when the channel is already open, blockade tracks neural activity — the busier the receptor, the more it is silenced.

PCP is not gentle about it. Ranked by affinity for the channel site and by antagonist potency, the order is MK-801 (dizocilpine) > PCP > ketamine: PCP blocks the NMDA receptor more potently than ketamine and, thanks to very different pharmacokinetics, for far longer (Lodge & Mercier, Br J Pharmacol 2015). That combination — high potency, long dwell — is why PCP’s effects are so much heavier and more protracted than a ketamine experience.

As with every psychoactive drug, the headline mechanism is not the whole chord. PCP also inhibits the dopamine transporter, acts at sigma receptors to modulate dopamine release, and antagonizes nicotinic acetylcholine receptors. The dopaminergic activity matters for the stimulant, euphoric, and psychotogenic coloring of the experience — but the NMDA blockade is the through-line that links PCP to its entire drug family and to its scientific importance.

The big idea: NMDA hypofunction and a model of psychosis

Here is where PCP stops being a footnote in anesthesiology and becomes central to psychiatry. The resemblance to psychosis was noticed immediately: the very first human study, Luby and colleagues in 1959, was titled “Study of a new schizophrenomimetic drug — Sernyl” (Luby et al., 1959). Something about blocking NMDA receptors did not just intoxicate people; it reproduced a recognizable illness.

Crucially, it reproduced the whole illness. Stimulants like amphetamine can mimic the positive symptoms of schizophrenia — paranoia, hallucinations — which underwrote the older dopamine hypothesis. But amphetamine psychosis stops there. PCP, as Javitt and Zukin argued in their landmark 1991 synthesis, uniquely reproduced the positive, negative, and cognitive dimensions together: emotional withdrawal and blunting, motor retardation, formal thought disorder, and the specific neuropsychological deficits that define the disorder (Javitt & Zukin, Am J Psychiatry 1991). Krystal and colleagues then confirmed in 1994 that even sub-anesthetic ketamine, given to healthy volunteers under controlled conditions, produced this broad schizophrenia-like picture (Krystal et al., Arch Gen Psychiatry 1994). No other pharmacological model came so close.

The inference was the NMDA-receptor hypofunction hypothesis: if too little NMDA signaling can manufacture the full syndrome in a healthy brain, then endogenous NMDA hypofunction may be part of what is schizophrenia. This did not overthrow the dopamine hypothesis so much as reframe it, giving the field a mechanism that could account for negative and cognitive symptoms — the ones dopamine drugs never explained and antipsychotics never treated well (Jentsch & Roth, Neuropsychopharmacology 1999). Phencyclidine, the failed anesthetic, had handed neuroscience a working hypothesis about one of medicine’s hardest illnesses.

The circuit: parvalbumin interneurons and cortical disinhibition

Why should blocking excitatory receptors produce a state of cortical overexcitation? The resolution is one of the more elegant ideas in modern circuit neuroscience. The NMDA receptors most sensitive to blockade by PCP and ketamine sit not on the main excitatory (pyramidal) neurons but on fast-spiking, parvalbumin-expressing GABA interneurons — the inhibitory cells that keep cortical output disciplined. Silence those interneurons and you release their brake. The pyramidal neurons they normally restrain fire unchecked, producing a paradoxical surge of glutamate and a burst of disorganized, poorly synchronized activity — measurable as disrupted gamma-band oscillations, the rhythm on which coherent perception and working memory depend (Nakazawa & Sapkota, Pharmacol Ther 2020).

This “cortical disinhibition” model connects a great deal at once. It explains the glutamate elevations seen with these drugs; it explains why parvalbumin-interneuron deficits are among the most reproducible findings in postmortem schizophrenia tissue; and it offers a route from a glutamate problem to the downstream dopamine dysregulation of the classic hypothesis. Repeated PCP administration in animals reliably lowers cortical GABA markers and disrupts prefrontal function, and it remains a workhorse model for testing candidate treatments (Amitai et al., Neuropharmacology 2012). The felt chaos of a PCP experience and the persistent architecture of psychosis, on this account, share a common mechanical root: the wrong cells went quiet.

Pharmacokinetics and the dose-response ladder

PCP’s clinical character is written in its pharmacokinetics. Smoked — the most common route, typically as leafy material soaked in the drug — effects begin in 2–5 minutes; swallowed, in 30–60 minutes. A typical intoxication lasts 4–8 hours, but subjective effects can linger 24–48 hours, and the reported elimination half-life is long and highly variable (commonly cited around 21 hours, with a wide range across individuals). PCP is metabolized largely by hepatic cytochrome P450 (CYP3A prominent) to inactive metabolites (Doyno & Ganti, StatPearls 2023).

The property that makes PCP clinically treacherous is its high lipophilicity: it partitions into fat and brain tissue and is released back into the blood slowly. This produces the phenomenon of “reintoxication” — fluctuating, recrudescent symptoms as stored drug re-enters circulation, sometimes flaring days after the last use. Because a weak base is excreted faster in acidic urine, urinary pH also shapes clearance.

Effects climb a rough dose ladder. Low-to-moderate doses bring detachment, numbness, slurred speech, and loss of coordination, often with a striking sense of strength and invulnerability, accompanied by the drug’s clinical signature — a blank stare, catatonic posturing, and nystagmus (rapid involuntary eye movements, classically vertical, horizontal, or rotary). Higher doses add frank hallucination, agitation, and psychosis; higher still, catatonia, seizures, coma, and life-threatening hyperthermia. It is the same molecule throughout — the dose sets the face it shows. A compounding real-world problem is that illicitly produced PCP varies widely in purity and is often sold soaked into plant material at unknown concentration, so the boundary between a low dose and a dangerous one is frequently invisible to the person taking it.

Safety and harms, held honestly

PCP carries a fearsome cultural reputation, and an evidence-forward account has to separate the documented risks from the folklore. The single best dataset remains McCarron’s 1981 series of 1,000 consecutive intoxications (McCarron et al., Ann Emerg Med 1981). It is sobering precisely because it is unsensational: only 57% of patients had the “classic” nystagmus-plus-hypertension picture, 46% were alert and oriented, and violence occurred in about 35% of cases — real and important, but far from the universal superhuman rampage of legend. The most common presentation was not raging aggression but a mixed, fluctuating picture of bizarre behavior, agitation, and disorientation, and a large share presented instead as catatonic, stuporous, or comatose.

The genuine dangers are largely secondary. Rhabdomyolysis — muscle breakdown that can cause acute kidney injury — appeared in roughly 2% of that series, with a few patients requiring dialysis; hyperthermia driven by agitation and muscular hyperactivity is a leading contributor to the rare deaths, alongside trauma (from impaired judgment, analgesia, and that false sense of invulnerability). This is why modern emergency management emphasizes a calm, low-stimulation environment and chemical sedation with benzodiazepines rather than physical restraint — restraint against a hyperthermic, hyperactive patient worsens rhabdomyolysis and heat injury. Prolonged psychosis outlasting the drug is well described, particularly in vulnerable individuals, consistent with everything the schizophrenia model would predict.

Chronic use has its own signature. Repeated PCP exposure is associated with dependence and a withdrawal syndrome, and with persistent problems that echo the acute pharmacology — impaired memory and thinking, speech difficulties, anxiety, depression, and social withdrawal that can outlast use. These lingering deficits are, again, exactly what the schizophrenia model would anticipate from sustained NMDA hypofunction. One research caveat deserves mention: in 1989 John Olney showed that high doses of NMDA antagonists — including PCP — produce reversible vacuole formation in specific rat cortical neurons (“Olney’s lesions”), with potency again tracking MK-801 > PCP > ketamine (Olney et al., Science 1989). The finding shaped decades of safety thinking about this drug class; its direct relevance to typical human exposures is debated, but it is part of an honest picture.

From PCP’s failure to ketamine’s cure

PCP itself has no therapeutic use — it is too potent, too long, too unpredictable, and too prone to psychosis. Its contribution to medicine is entirely intellectual, and it is large. The chain runs directly from the 1983 discovery that dissociatives are NMDA antagonists, through the hypofunction model of psychosis, to the recognition that the NMDA receptor is a druggable lever on serious mental illness. That recognition is exactly what made researchers ask whether NMDA modulation might help depression — and the answer, delivered by ketamine, was one of the most important findings in modern psychiatry: a single sub-anesthetic dose can lift treatment-resistant depression within hours. The glutamate hypothesis of depression, and the 2019 FDA approval of esketamine (Spravato), are downstream of insights that began with a discarded anesthetic (Zanos & Gould, Phil Trans R Soc B 2024). The current frontier — subunit-selective and circuit-targeted approaches meant to keep the antidepressant benefit while shedding the dissociation — is the same story still being written. PCP was the tool that found the door; ketamine is what walked through it.

The honest bottom line

Phencyclidine is a rare case of a drug whose scientific value vastly exceeds — and inverts — its clinical value. On the street it is a heavy, long, and genuinely hazardous dissociative; in the clinic it is nothing, having failed the one job it was designed for. In the laboratory it is foundational: the compound that revealed the PCP site, anchored the NMDA-hypofunction model of schizophrenia, exposed the parvalbumin-interneuron circuit, and lit the path to ketamine. The limitations are real and worth stating plainly — the hypofunction model is a powerful framework, not a complete theory of schizophrenia; the parvalbumin-disinhibition account is well-supported but still being refined; and much of the mechanistic detail rests on animal work whose mapping onto human illness is imperfect. Holding all of that together — a dangerous drug, a useless medicine, and an indispensable scientific instrument — is the only accurate way to think about PCP.

OOTW Journal is educational and does not provide medical advice. PCP has no accepted medical use and is a Schedule II controlled substance in the United States; this article is not a guide to obtaining or using it. Acute PCP intoxication is a medical emergency — agitation, high fever, muscle rigidity, or altered consciousness after suspected use warrants emergency care or a call to Poison Control (1-800-222-1222 in the US). This article is education, not medical advice.