L-Tetrahydropalmatine

L-THP / Corydalis extract

L-Tetrahydropalmatine (L-THP / Corydalis extract) is a cognitive & mood research compound. Dopamine D1 and D2 receptor antagonist from Corydalis yanhusuo, used in traditional Chinese medicine for pain and sleep. The analgesia appears to be dopaminergic rather than opioid.

How L-Tetrahydropalmatine works

Dopamine D1 and D2 receptor antagonist from Corydalis yanhusuo, used in traditional Chinese medicine for pain and sleep. The analgesia appears to be dopaminergic rather than opioid.

Proposed benefits

Researched for focus, memory, neuroprotection, mood and stress resilience.

Where to get L-Tetrahydropalmatine

The evidence for L-Tetrahydropalmatine

Graded by what exists behind each claim.

✅ Clinically validated

• Used clinically in China as an analgesic and sedative, with Chinese clinical literature behind it. Small Western trials have examined it for cocaine dependence, reporting reduced use — a genuinely interesting application that has not been pursued at scale.

📊 Correlative data

• Long traditional use as a component of *Corydalis yanhusuo* in Chinese medicine, and modern use for pain and sleep. Hepatotoxicity case reports exist, mostly linked to adulterated or mislabelled products rather than to the isolated alkaloid.

🧪 Theoretical / extrapolated

• An isoquinoline alkaloid acting as a dopamine D1 and D2 antagonist with additional GABAergic activity.
• Dopamine antagonism is an unusual mechanism for something sold as a wellness compound, and it predicts the risks that framing hides: sedation, and the theoretical possibility of extrapyramidal effects with sustained use, the same class of effect antipsychotics cause.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What L-Tetrahydropalmatine actually does

The single most useful sentence about this alkaloid is that it is pharmacologically a neuroleptic sold as a herbal sleep aid. Marcenac 1986 characterized l-tetrahydropalmatine in rat striatum as a dopaminergic antagonist that can block both post- and presynaptic receptors. Not a modulator, not an adaptogen — a dopamine receptor blocker, which is the same pharmacological class as haloperidol and metoclopramide.

And blocking both sides of the synapse produces a paradox that is worth working through, because it explains the whole felt experience. The presynaptic D2 autoreceptor is the brake a dopamine neuron uses to sense its own output and slow down. Block it and the brake comes off: the neuron fires harder and releases more. Marcenac 1986 measured exactly that — extracellular dopamine rose to 220% of the basal value. But the postsynaptic receptors are blocked at the same time, so the extra dopamine has nowhere to act, and it gets metabolized instead: extracellular DOPAC rose to 155 ± 9% of control, and post-mortem tissue DOPAC rose 250% while total dopamine content was unchanged. High turnover, blocked signaling. That combination — the neuron shouting into a closed receptor — is the classic neurochemical signature of a dopamine antagonist, and it is why the sedation is reliable and why the withdrawal question is a real one.

The analgesia claim, read honestly. l-THP’s pain effect is genuinely non-opioid, which is what makes it interesting. But “dopaminergic rather than opioid” means the analgesia belongs to the same mechanism class as a neuroleptic’s, and dopamine antagonists have been used for pain for exactly that reason for sixty years. It is a real mechanism with a real cost.

There is a second and a third arm, both from the psychostimulant work. Yun 2014 found l-THP inhibited methamphetamine-induced hyperlocomotion in rats at 10 and 15 mg/kg, attenuated the methamphetamine-induced fall in dopamine D3 receptor mRNA, and enhanced 5-HT neuronal activity. So the pharmacology is not confined to D1/D2: there is a D3 component and a serotonergic one. That breadth is why it shows up in addiction research and why a single-receptor account of it is incomplete.

Cell, rodent, human — and where it stops

Step one, rat brain, in vivo microdialysis and post-mortem tissue. Marcenac 1986 gave rats l-THP at a low dose of 1 mg/kg intraperitoneally and at sedating doses of 5–10 mg/kg, and measured striatal dopamine and its metabolite: DOPAC up 250% post-mortem with unchanged dopamine content, extracellular DOPAC at 155% and extracellular dopamine at 220% of baseline, and the conclusion of pre- and postsynaptic antagonism.

Step two, rats and mice, a behavioral model. Yun 2014 used 10 and 15 mg/kg and showed significant inhibition of methamphetamine-induced hyperlocomotion, with the D3 mRNA and 5-HT findings above. Two rodent species, consistent direction.

Step three, humans — and there is a proper controlled study, which is more than most of this catalog can say. Hassan 2017 ran a randomized, double-blind, placebo-controlled pharmacokinetic and safety study in cocaine users: 24 enrolled, 20 completed, 19 provided pharmacokinetic data, taking l-THP 30 mg twice a day orally for 4 days. It reported no significant differences in adverse effects or vital signs against placebo, and an interaction result: co-administered cocaine AUC was 211.5 against 261.4 h·ng/mL and Cmax 83.3 against 104.5 ng/mL on l-THP versus placebo — roughly a fifth lower on both.

The obstacle, stated exactly, and it has two parts. First, the human study that exists was designed to answer a safety and interaction question over four days in people who use cocaine Hassan 2017. It is not an efficacy trial for pain or for sleep, and no efficacy endpoint was measured. The Chinese clinical literature that underpins the traditional use is not indexed in a form that can be checked here, and a claim that cannot be checked cannot carry a citation on this site. Second, the dose: converting the rodent sedating doses by body-surface-area factors (rat 6, human 37) puts 5–15 mg/kg in a rat at roughly 57 to 170 mg for a 70 kg adult. The 100–200 mg doses this site records therefore sit inside or above the scaled range that produced frank sedation and dopamine antagonism in rats — and more than three times the 60 mg per day used in the only controlled human study. Unlike most compounds in this cohort, the everyday dose here is not conservatively low.

L-Tetrahydropalmatine pharmacokinetics — how much of it actually gets in

Route: oral, which is both the traditional route and the one used in the controlled study — 30 mg twice daily for 4 days Hassan 2017. There is no injectable or intranasal l-THP with published human pharmacokinetics.

What is actually published about its own disposition, and what is not. The only controlled human pharmacokinetic study of l-THP reported the parameters of the co-administered drug rather than of l-THP itself: cocaine AUC 211.5 vs 261.4 h·ng/mL and Cmax 83.3 vs 104.5 ng/mL Hassan 2017. So there is no citable human half-life, no bioavailability figure and no clearance route for the alkaloid, and this page will not manufacture one.

But the interaction result is itself a pharmacokinetic finding, and it is the most practically important one on the page. Four days of 60 mg a day lowered a co-administered drug’s peak concentration by about 20% and its total exposure by about 19% Hassan 2017. Whatever the mechanism — slowed gastric emptying, which dopamine antagonists reliably cause, or an enzyme effect — the conclusion is the same: l-THP is not pharmacokinetically inert. It changed the exposure of another drug in four days at a dose one-third of what people take. Anything else taken alongside it should be assumed to be affected until somebody shows otherwise.

What degrades it: an unanswered question with a well-defined shape. l-THP is a tetrahydroprotoberberine isoquinoline alkaloid, the structural class that is characteristically cleared by hepatic oxidation and conjugation, and no human cytochrome assignment has been published for it. The hepatic case series Woolf 1994 is indirect evidence that the liver handles it: injury appeared after a mean of 20 weeks and resolved after stopping over a mean of 8 weeks, which is the time course of a metabolism-linked idiosyncratic injury rather than of a direct dose-dependent toxin.

The oral barrier is mild, and that is part of the problem. This is a small, lipophilic, uncharged tertiary amine — the physical profile that crosses membranes easily, including the blood-brain barrier, which is why an oral dose produces central effects at all. Good absorption plus a dopamine-antagonist mechanism plus a dose range that overlaps the scaled rodent sedating dose is a combination that makes the compound easy to overshoot.

What would have to be true, and how you would know it was not

Four predictions. The first is the confirmation nobody has ever run, the third is the one that cuts against the way this is sold, and the fourth has a published timescale attached.

1. Prolactin should rise, and this is the single cleanest test on the page. Dopamine is the physiological brake on pituitary lactotrophs, delivered through D2 receptors that sit outside the blood-brain barrier. A dopamine antagonist releases that brake and prolactin climbs — it is how every antipsychotic’s D2 occupancy is confirmed in practice. Draw prolactin at a fixed morning hour before starting and again after two to four weeks of regular use. Marcenac 1986 establishes the antagonism; nobody has ever measured the prolactin consequence in a human taking l-THP. If prolactin does not move at a sedating dose, the antagonist account of this compound is wrong in people.

2. Liver enzymes, on the timescale the case series defines. Run a Comprehensive Metabolic Panel (CMP) and a GGT at baseline, at 6 weeks and at 12 weeks. That schedule is not arbitrary: hepatitis in Woolf 1994 appeared after a mean of 20 weeks, range 7 to 52, so a single check at 4 weeks is early enough to be reassuring and too early to be informative. This is the one compound in this cohort where the monitoring interval comes straight out of the published injury data.

3. The prediction that cuts against the framing: stopping abruptly after sustained use should be worse than stopping a sedative. The mechanism leaves the presynaptic brake off and the postsynaptic receptor blocked — extracellular dopamine at 220% of basal Marcenac 1986. Remove the blockade and that elevated release meets unblocked receptors. The falsifiable prediction is a rebound on withdrawal — agitation, insomnia worse than baseline, restlessness — appearing within a day or two of stopping after weeks of use, and it is the opposite of what a “gentle herbal” framing predicts. Test it with a 14-night baseline, several weeks on, then a deliberate 7-night off block, scored the same way throughout.

4. And the movement prediction, which is the one to watch for. Sustained dopamine antagonism is the setting in which extrapyramidal effects appear — a fine resting tremor, stiffness, and akathisia, which is felt as an inability to sit still rather than as anxiety. The dose arithmetic above says 100–200 mg sits in the scaled range of the rodent sedating doses, so this is not a remote theoretical concern at the doses in use. It has never been looked for prospectively with a rating scale in anybody taking l-THP, which is why it belongs in the predictions rather than in the evidence.

What nobody has tested yet

Nobody has measured prolactin on l-THP in a human. The compound is a characterized dopamine antagonist Marcenac 1986, prolactin is the standard peripheral read-out of D2 blockade, and the test costs almost nothing. A dozen people with before-and-after draws would establish whether this alkaloid produces measurable D2 occupancy in a person at the doses actually taken — which is the question every other claim on this page depends on.

Nobody has published l-THP’s own human pharmacokinetics. The one controlled human study reported the co-administered drug’s parameters Hassan 2017. Without a half-life there is no rational dosing interval, and for a compound with a plausible rebound on withdrawal the interval is exactly what matters.

Nobody has separated the alkaloid from the product in the hepatotoxicity question, and this is the most consequential open question here. The seven hepatitis cases Woolf 1994 and the six-case series Horowitz 1996 both concern Jin Bu Huan, a manufactured herbal preparation. A purified single alkaloid has never been through a controlled hepatic-safety study of meaningful duration, and the one controlled study that exists ran for four days Hassan 2017. Whether the liver injury belongs to l-THP or to something else in the tablet decides whether a purified product is safer, and nothing published answers it.

And nobody has tested it against a sedative-hypnotic on sleep architecture. A dopamine antagonist produces sedation; a hypnotic produces sleep. Those are not the same thing and a polysomnograph distinguishes them in one night. Given that this compound is used primarily for sleep, the absence of a single published sleep-architecture study is the most surprising gap in its literature.

L-Tetrahydropalmatine — its own safety story, not its class's

The hepatitis series is specific, published, and belongs on this page rather than in any class block. Woolf 1994 described seven patients with acute hepatitis after a mean of 20 weeks (range 7 to 52) of Jin Bu Huan use. The biopsy findings matter more than the enzyme numbers: one case showed hepatitis with eosinophils — the signature of an immune-mediated, idiosyncratic reaction rather than a dose-dependent toxin — and another showed moderate fibrosis and microvesicular steatosis, which points at mitochondrial injury. Six of the seven resolved after stopping, over a mean of 8 weeks (range 2 to 30). Reversible, but slowly, and only if you stop.

The pediatric finding is the one that changes where you keep it. Horowitz 1996 reported six cases: three children and three adults. The adults developed hepatitis from long-term use. The children, after acute ingestion, had life-threatening neurologic and cardiovascular manifestations — and the series names the active constituent as levo-tetrahydropalmatine, a potent neuroactive substance. A tablet that looks like a supplement, in a house with a child, is the exact scenario in that paper.

The dopamine-antagonist risk profile, which the herbal framing hides. Sedation is the intended effect and it is the mildest consequence. The class also produces hyperprolactinemia — which in a man presents as reduced libido and in a woman as cycle disturbance, and which nobody measures because nobody thinks of this as a dopamine drug — and, with sustained exposure, extrapyramidal movement effects. Marcenac 1986 establishes the receptor pharmacology that predicts all of it, and no study of l-THP has looked for any of it in a person.

What is genuinely reassuring, and the exact limits of it. Hassan 2017 found no significant differences in adverse effects or vital signs against placebo. That is real, controlled, human safety data and this page will not talk it down. It covers 60 mg a day, for four days, in 20 people. It does not cover 100–200 mg, it does not cover weeks, and the hepatotoxicity in the case literature appeared after a mean of twenty weeks Woolf 1994. The reassurance and the harm reports describe different doses and different durations, and both are true.

Sources read for this page

• Marcenac F. Effect of l-tetrahydropalmatine on dopamine release and metabolism in the rat striatum.. Psychopharmacology (Berl) 1986 · PMID 3090599
• Yun J. L-tetrahydropalmatine inhibits methamphetamine-induced locomotor activity via regulation of 5-HT neuronal activity and dopamine D3 receptor expression.. Phytomedicine 2014 · PMID 25172791
• Hassan HE. Pharmacokinetics and Safety Assessment of l-Tetrahydropalmatine in Cocaine Users: A Randomized, Double-Blind, Placebo-Controlled Study.. J Clin Pharmacol 2017 · PMID 27363313
• Woolf GM. Acute hepatitis associated with the Chinese herbal product jin bu huan.. Ann Intern Med 1994 · PMID 7944049
• Horowitz RS. The clinical spectrum of Jin Bu Huan toxicity.. Arch Intern Med 1996 · PMID 8774209

L-Tetrahydropalmatine — safety, predicted from mechanism

L-Tetrahydropalmatine — interference & stacking

Bloodwork to run alongside L-Tetrahydropalmatine

Baseline first, then again at 8–12 weeks.

The Brain Fog & Cognition panel covers these in one order — 12 markers, $233.06 with the discount applied.

Check results you already have → · All 103 markers A–Z

L-Tetrahydropalmatine — frequently asked questions

What L-Tetrahydropalmatine is used for

L-Tetrahydropalmatine appears under 1 goal in the goal router.

Related Cognitive & Mood compounds

← Explore the full Protocol Vault

↑ Back to on this page