Pramiracetam

Fat-soluble racetam

Pramiracetam (Fat-soluble racetam) is a cognitive & mood research compound. Potent racetam that enhances high-affinity choline uptake in the hippocampus — strong memory effect.

How Pramiracetam works

Potent racetam that enhances high-affinity choline uptake in the hippocampus — strong memory effect.

Proposed benefits

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

Where to get Pramiracetam

The evidence for Pramiracetam

Graded by what exists behind each claim.

✅ Clinically validated

• Small human trials exist, including work in traumatic brain injury patients reporting improved memory. The studies are decades old, small, and were not followed by a development program — which is the pattern across the whole racetam family.

📊 Correlative data

• Community use focused on memory and long-form work. Reported as the most potent of the classical racetams by weight, and strictly fat-soluble.

🧪 Theoretical / extrapolated

• Increases high-affinity choline uptake in the hippocampus — a more specific cholinergic mechanism than the other racetams, which act more broadly on AMPA signaling.
• That specificity predicts the practical rule: it depends on available choline, so it does less in someone whose intake is poor, and headache without a choline source is the mechanism running short of substrate.

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 Pramiracetam actually does

The molecule is piracetam with one thing bolted on, and that one thing is the whole compound. Piracetam is 2-oxo-1-pyrrolidineacetamide, C6H10N2O2, 142.16 Da — small, very polar, dosed in grams. Pramiracetam keeps that core and replaces the primary amide with a bis(1-methylethyl)aminoethyl amide: C14H27N3O2, 269.38 Da, sold as the sulfate salt (CI-879). Adding a bulky, basic, branched alkylamine turns a water-soluble molecule into a lipophilic one, and that is the mechanical reason the dose falls from grams to hundreds of milligrams and the reason the catalog says to take it with fat.

The target is a transporter with a gene name, which is rare in this class. Pramiracetam's described action is enhancement of sodium-dependent high-affinity choline uptake in hippocampus. That is CHT1, the product of SLC5A7, and it is the rate-limiting step of acetylcholine synthesis: choline taken up by CHT1 is the choline that choline acetyltransferase turns into transmitter. Every other racetam is described in terms of AMPA-receptor modulation or membrane fluidity, which are hard to test; this one names the single enzymatic bottleneck of a neurotransmitter system.

What CHT1 biology predicts, and nobody has written this down. CHT1 is not sitting statically in the presynaptic membrane. Most of it is held in the membrane of synaptic vesicles and inserted into the plasma membrane when those vesicles fuse — so high-affinity choline uptake rises with the firing rate of the cholinergic terminal. A drug that increases the efficiency of that transporter therefore has more to work with in an active circuit and nothing to work with in a quiet one. That is a mechanistic prediction of exactly the subjective report this compound generates: people describe it as doing nothing while idle and being noticeable during sustained work. It is the difference between a drug that adds transmitter and a drug that raises the ceiling on demand-driven synthesis.

The human experiment that tests the mechanism directly. Scopolamine is a muscarinic antagonist; blocking postsynaptic muscarinic receptors produces a reliable, reversible amnesia in healthy people. Mauri 1994 gave 600 mg twice daily for 10 days, then 0.5 mg of scopolamine hydrobromide intramuscularly, and pramiracetam partially reduced the amnesia in both a young group (18–42) and an older one (55–65), 12 males in each Mauri 1994. A drug that partially rescues muscarinic blockade by raising presynaptic acetylcholine synthesis is mechanistically coherent. A drug acting only on AMPA receptors would not be expected to.

And the shape of the dose-response, which is the single most practically important sentence in this compound's literature. The opening line of the Alzheimer's trial states that pramiracetam's cognitive-enhancing effects in animal models of learning and memory are characterized by an inverted U-shaped dose-response curve Claus 1991. Not ‘more is better’, not ‘more is worse’ — more is worse past a peak nobody has located in a person. That sentence is on no vendor page anywhere.

Cell, rodent, human — and where it stops

The animal rung is real and largely invisible to PubMed. The high-affinity choline uptake work that this compound's entire mechanism rests on was published in the early 1980s, some of it in journals that PubMed does not index, which is why a search for ‘pramiracetam choline uptake’ returns almost nothing. That is a fact about the evidence rather than about the drug, and it is the reason the mechanism section above is written from transporter biology and a human challenge study instead of from the primary rodent papers.

Human study one: the mechanism challenge. 24 healthy males in two age bands, 1,200 mg/day for 10 days, scopolamine 0.5 mg intramuscularly on day 11, psychometrics before and at 1, 3 and 6 hours. Scopolamine impaired episodic memory and selective attention; visuo-motor and incidental learning were untouched; pramiracetam partially reduced the amnesia in both age groups Mauri 1994.

Human study two: the clinical result, and it is the strongest in the racetam class. Young males with memory and cognitive problems after head injury or anoxia, double-blind and placebo-controlled, 400 mg of pramiracetam sulfate three times a day. Memory improved — especially delayed recall — to a degree the authors called clinically significant, the improvement was maintained across an 18-month open trial, and it was still present through a 1-month follow-up after the drug was stopped McLean 1991.

Human study three: the failure, and it is the more instructive one. 10 patients with probable Alzheimer's disease, a two-phase enrichment design built specifically to accommodate an inverted U by finding each person's own best dose. 8 of 10 found a best dose in phase one; only 2 reproduced it in the replication phase. Fluorodeoxyglucose PET in two patients showed no definite change, and the conclusion was that doses up to 4,000 mg are unlikely to help in Alzheimer's disease Claus 1991. Set beside the Cochrane verdict on the parent compound — the published evidence does not support piracetam in dementia or cognitive impairment, with an effect on global impression of change and no benefit on any specific measure Flicker 2001 — the class pattern is clear.

The obstacles, and the first is a dose mismatch nobody flags. (1) Every human result used 1,200 mg/day. This site lists 300–600 mg. That is a quarter to a half of the only dose that has ever produced a human signal, and on an inverted U a lower dose is not simply a weaker version of the same thing — it can sit on the flat part of the curve and be indistinguishable from nothing. (2) The persistence result cannot be interpreted. McLean's improvement lasted a month past discontinuation McLean 1991, which is either a genuine durable effect on recovery after brain injury or a practice effect from repeated memory testing — and because that observation was made in the open extension rather than the blinded phase, the study cannot tell them apart. (3) All three human studies are in males. There is no female pramiracetam data of any kind, which is unusual even by the standards of 1991. (4) The trials ran in impaired brains. Post-traumatic amnesia, anoxic injury, scopolamine blockade, Alzheimer's: every positive signal is a rescue of a degraded cholinergic system. Nothing has ever been measured in an intact one.

Pramiracetam pharmacokinetics — how much of it actually gets in

The catalog says ~5 hours, and the trials confirm it indirectly: every human study dosed two or three times a day. 400 mg three times daily in the head-injury trial McLean 1991, 600 mg twice daily in the scopolamine study Mauri 1994. A once-daily protocol has never been tested and would not fit the half-life.

What degrades it. The linkage is a secondary amide, not an ester, so plasma carboxylesterases do not open it the way they open the ethyl ester nootropics — and there is no peptide bond for aminopeptidases or brush-border peptidases to attack either. The disposal route is hepatic: oxidative N-dealkylation of the branched diisopropylaminoethyl side chain, then renal excretion. Practically that means the pharmacokinetic variable most likely to differ between two people is liver oxidative capacity, not gut enzymatic hydrolysis.

The oral barrier, and the food effect nobody has measured. The lipophilicity that the diisopropylaminoethyl group confers is the reason for the ‘take with fat’ instruction, and the mechanism is ordinary: dietary lipid slows gastric emptying, stimulates bile, and improves dissolution of a poorly water-soluble base. The consequence is that a fasted 400 mg dose and a fed 400 mg dose are not the same exposure, and no fed/fasted AUC ratio has ever been published for this compound. That is the standard regulatory food-effect study — 12 subjects, two visits, one assay — and its absence means nobody can say whether the site's 300–600 mg taken with a meal is closer to or further from the trials' 1,200 mg taken under unspecified conditions.

The substrate arithmetic, which is the part that actually changes practice. The mechanism consumes choline: raising high-affinity choline uptake without raising choline supply moves the bottleneck one step upstream, which is the mechanistic account of the headache this compound is famous for. The US adequate intake for choline is 550 mg/day for men and 425 mg/day for women, and population intakes commonly sit below both. Two whole eggs supply roughly 300 mg. So the practical statement is not ‘take a choline source’ but ‘your baseline choline intake, in milligrams, is a variable in this compound's dose-response’, and it is a variable no trial has ever recorded.

Route. Oral only, in every human study and every protocol.

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

Four predictions. The first argues against the dose almost everyone uses.

1. Against: 300 mg/day will do nothing measurable, and that is a prediction about the dose rather than about the drug. Every human signal came from 1,200 mg/day McLean 1991 Mauri 1994, and the dose-response is an inverted U mapped only in animals Claus 1991. Prediction: a person running 300 mg and reporting a clear effect is describing expectancy, and a person running 1,200 mg is running the only regimen with evidence and also the only one with an unmeasured distance from the top of the curve. This is the rare case where the honest recommendation is to match the trial rather than to start low, and the rare case where nobody can tell you what the ceiling is.

2. Trail Making and MoCA will not move in an intact adult. Both instruments ceiling in healthy people, and every positive human result for this compound came from a damaged cholinergic system — head injury, anoxia, pharmacological blockade McLean 1991 Mauri 1994. Retest at 4 weeks if you like; understand that a null is uninformative and that the absence of a sensitive at-home test is why this compound's community reputation is so much more confident than its data.

3. The headache is falsifiable, and the test is a crossover on choline intake rather than on dose. If the mechanism is substrate depletion downstream of enhanced high-affinity choline uptake, then headache frequency should track choline intake inversely at a fixed pramiracetam dose, and should be near-absent in somebody already eating 550 mg/day. Protocol: two weeks at a fixed dose on a low-choline diet, two weeks at the same dose with two eggs a day added, headache days counted. If the headache does not respond to choline, the substrate story is wrong and the effect is something else — which would be a genuinely new finding about the most-repeated claim in the racetam community.

4. Homocysteine is the control, and it should not move. It is on the list because the usual answer to the headache is a choline source, and choline is oxidized to betaine, which donates a methyl group to homocysteine and lowers it. So a homocysteine change on this protocol would be evidence about the choline you added rather than about the pramiracetam, and confusing the two is exactly the error a stacked protocol invites. Draw it at baseline and 12 weeks, and hold B12, folate and B6 constant across the window.

What nobody has tested yet

Five experiments. The second one is the missing number that would let anyone dose this rationally.

1. There is no food-effect study. A compound whose own dosing instruction is ‘take with fat’ has never had a fed/fasted pharmacokinetic comparison published. Twelve subjects, two visits, one LC-MS assay. Until it exists, the milligram numbers on every page including this one describe a dose rather than an exposure.

2. Nobody has mapped the inverted U in a human. The curve is asserted from animal models Claus 1991 and the enrichment trial built to exploit it failed to replicate individual best doses Claus 1991. A four-arm dose-ranging study — 300, 600, 1,200 and 2,400 mg/day — with a sustained-attention task in healthy adults would locate the peak, and it is the only experiment that could make a dosing recommendation for this compound non-arbitrary.

3. There is no female data. Three human studies, all male McLean 1991 Mauri 1994 Claus 1991. Given that hepatic oxidative clearance and body composition both differ, the assumption that the same milligram dose produces the same exposure is untested in half the population.

4. Nobody has run the choline crossover. ‘Choline is a must’ is the single most repeated piece of advice about this compound and there is no experiment behind it. Pramiracetam with and without 500 mg of alpha-GPC, crossover, headache days and delayed recall as endpoints, 20 people. It would confirm or destroy the mechanistic story in six weeks.

5. Nobody has re-run McLean. The best human result in the whole racetam class is a small 1991 study in young men after head injury McLean 1991, never replicated, in an era before modern neuropsychological batteries or diffusion imaging. A modern repeat with an imaging endpoint would either establish the first genuinely useful indication for a racetam or close a thirty-five-year-old question.

Pramiracetam — its own safety story, not its class's

The class block worries about stimulants and liver enzymes. Pramiracetam is neither a stimulant nor a hepatically-activated prodrug, and its own story is these four things.

1. The interesting interaction is cancellation, not danger. A drug whose mechanism is raising presynaptic acetylcholine synthesis works directly against every anticholinergic a person is taking — older sedating antihistamines, some bladder antimuscarinics, tricyclics, some antipsychotics. That is exactly the axis Mauri's scopolamine study sits on Mauri 1994. The practical consequence is not toxicity; it is that somebody on a daily anticholinergic is running two drugs that partly erase each other, and neither prescriber knows.

2. The headache has a mechanism, so it has a fix that is testable. It is the most consistently reported effect of this compound and, on the substrate account above, it should respond to choline intake and should be dose-dependent. If it does neither in a given person, that person should stop rather than escalate, because the working explanation has failed for them.

3. What 1,200 mg/day for 18 months actually established. McLean's open extension ran that dose for a year and a half without reporting a safety signal McLean 1991. That is a genuine and unusually long human exposure record for anything in this catalog, and it is worth more than the absence of reports elsewhere. It is also a single small cohort of young men with brain injuries, monitored, in 1991 — not a general safety statement.

4. The supply, and here the dose size is the protection. Analysis of cognitive-enhancement supplements found 75% of declared quantities inaccurate and exposures up to four-fold above pharmacologic doses Cohen 2021; European and Australian medicines control laboratories documented 34 distinct unauthorized molecules across 159 samples and intercepted racetam-family compounds — phenylpiracetam among them — as bulk raw material Vanhee 2025. For a compound dosed at hundreds of milligrams with an 18-month human exposure record, a four-fold error lands inside a range that has been given to people — which is not true of the milligram-dosed compounds elsewhere in this cohort, and is the one place pramiracetam's unfashionably large dose is an advantage.

Sources read for this page

• McLean A Jr, Cardenas DD, Burgess D, Gamzu E. Placebo-controlled study of pramiracetam in young males with memory and cognitive problems resulting from head injury and anoxia. Brain Injury 1991;5(4):375-380 · PMID 1786500
• Mauri M, Sinforiani E, Reverberi F, Merlo P, Bono G. Pramiracetam effects on scopolamine-induced amnesia in healthy volunteers. Archives of Gerontology and Geriatrics 1994;18(2):133-139 · PMID 15374306
• Claus JJ, Ludwig C, Mohr E, Giuffra M, Blin J, Chase TN. Nootropic drugs in Alzheimer's disease: symptomatic treatment with pramiracetam. Neurology 1991;41(4):570-574 · PMID 2011259
• Flicker L, Grimley Evans G. Piracetam for dementia or cognitive impairment. Cochrane Database of Systematic Reviews 2001;(2):CD001011 · PMID 11405971
• Cohen PA, Avula B, Wang YH, Zakharevich I, Khan I. Five Unapproved Drugs Found in Cognitive Enhancement Supplements. Neurology: Clinical Practice 2021;11(3):e303-e307 · PMID 34484905
• Vanhee C, Deconinck E, George M, Hansen A, Hackl A, Wollein U, El-Atma O, Beerbaum N, Aureli F, Borioni A, et al. The Occurrence of Illicit Smart Drugs or Nootropics in Europe and Australia and Their Associated Dangers: Results from a Market Surveillance Study by 12 Official Medicines Control Laboratories. Journal of Xenobiotics 2025;15(3):88 · PMID 40558871

Pramiracetam — safety, predicted from mechanism

Pramiracetam — interference & stacking

Bloodwork to run alongside Pramiracetam

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

Pramiracetam — frequently asked questions

What Pramiracetam is used for

Pramiracetam appears under 1 goal in the goal router.

Related Cognitive & Mood compounds

← Explore the full Protocol Vault

↑ Back to on this page