Phenibut

GABA_B Agonism + α2δ Calcium-Channel Ligand

GABA itself will not cross the blood–brain barrier — it is too polar. Phenibut is the medicinal-chemistry trick that fixes that: a phenyl ring on the β-carbon makes the molecule lipophilic enough (and gives it a transporter handle) to reach the CNS, where the intact GABA backbone lets it act at GABA_B receptors. That is the same design as baclofen — phenibut is baclofen with the para-chlorine removed — and it explains why the two share a pharmacology and a withdrawal syndrome.

But GABA_B agonism is not the whole story, and probably not even the dominant one at real-world doses. Phenibut is racemic, and its active enantiomer R-phenibut binds the α2δ subunit of voltage-gated calcium channels — the gabapentin/pregabalin target — with ~4× higher affinity than it binds GABA_B (Zvejniece et al., 2015). So phenibut is genuinely a two-mechanism drug: a weak metabotropic GABA_B agonist stacked on a gabapentinoid. Both mechanisms are low-potency — this is a drug dosed in hundreds of milligrams to grams, not the nanomolar micrograms of a benzodiazepine or an opioid.

The net effect is broad, slow CNS depression with an anxiolytic bias — subjectively somewhere between alcohol, a low dose of GHB, and pregabalin, arriving hours late and staying most of a day. The low potency is deceptive: because a "dose" is a scoop of powder rather than a single small pill, the gap between an anxiolytic dose and an over-sedating one is measured by an easily-mis-scooped few hundred milligrams.

Slow In, Slow Out — and No Off-Ramp Metabolites

Phenibut's pharmacokinetics are the source of most of its trouble. It is a small, water-soluble zwitterionic amino acid, so it does not flood across the blood–brain barrier the way a lipophilic benzodiazepine does — it enters the CNS slowly, and peak effects lag ingestion by 2–4 hours. The subjective effects then persist for the better part of a day. Unlike diazepam, there is no cascade of active metabolites: phenibut undergoes little hepatic transformation and is largely excreted unchanged by the kidneys. The long tail is the parent drug's own slow clearance and slow CNS washout, not a metabolite chain.

Metabolism — the short version: what goes in is mostly what comes out. Phenibut is not activated or extended by metabolism the way a prodrug or diazepam is; the "long duration" is pharmacokinetic inertia, not a metabolite relay. That is also why hepatic enzyme interactions matter far less here than for benzodiazepines — and why renal impairment, not liver disease, is the accumulation risk.

The clinical trap is baked into these numbers. A user takes a dose, feels little at the 60–90 minute mark because the drug hasn't arrived yet, and redoses "because it isn't working" — then both doses land together three hours later on top of a drug that lingers for a day. Combined with rapidly building tolerance, this is the engine of dose creep: yesterday's gram becomes today's baseline, and the therapeutic window narrows with every escalation.

A Genuinely Nasty Discontinuation Syndrome

In the former Soviet bloc phenibut is a legitimate prescription drug for anxiety, insomnia, PTSD, stuttering, vestibular disorders, and alcohol-withdrawal support, and marketed as a mild "nootropic." Used occasionally at modest doses it is a fairly gentle anxiolytic. The problem is that almost none of the grey-market population uses it that way — and the pharmacology punishes daily use hard. What follows is not moralising; it is the predictable neuroadaptation of an agonist at an inhibitory system.

The central asymmetry is familiar from diazepam but sharper: an occasional, measured dose is one of the milder anxiolytics around, yet the drug builds dependence quickly and its withdrawal — not its acute toxicity — is where the real danger lives. Layer on an unregulated supply, eyeballed gram-scale dosing, and a slow onset that invites redosing, and phenibut becomes a drug that is easy to start, easy to escalate, and genuinely hard to stop.