Nefiracetam

DM-9384

Nefiracetam (DM-9384) is a cognitive & mood research compound. Racetam that potentiates both nicotinic and NMDA receptor currents and enhances GABAergic transmission — a broader receptor profile than piracetam.

How Nefiracetam works

Racetam that potentiates both nicotinic and NMDA receptor currents and enhances GABAergic transmission — a broader receptor profile than piracetam.

Proposed benefits

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

Where to get Nefiracetam

The evidence for Nefiracetam

Graded by what exists behind each claim.

✅ Clinically validated

• Reached phase 2 in humans for post-stroke apathy and depression. Results were equivocal and development was discontinued; a separate program in Alzheimer's did not proceed either.
• Rodent toxicology showed testicular effects at high doses, which is part of why development stopped and is worth knowing.

📊 Correlative data

• Limited community use relative to the other racetams. Reported experience is of a mood-lifting effect more than a cognitive one, consistent with the apathy indication it was trialed for.

🧪 Theoretical / extrapolated

• Potentiates GABA-A and nicotinic acetylcholine receptors and enhances calcium channel currents — a broader receptor profile than the AMPA-focused racetams.
• The GABAergic component is the likely explanation for the mood effect and distinguishes it mechanistically from the rest of the family.

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 Nefiracetam actually does

Nefiracetam is N-(2,6-dimethylphenyl)-2-(2-oxopyrrolidin-1-yl)acetamide: C14H18N2O2, molecular weight 246.30, XLogP 1.4, topological polar surface area 49.4 Ų National Center for Biotechnology Information 2026. Structurally it is piracetam — the 2-oxopyrrolidine acetamide core is identical — with the terminal amide nitrogen carrying a 2,6-dimethylphenyl group instead of two hydrogens. That single aryl substitution is why this molecule is dosed at 100–300 mg where piracetam is dosed in grams: it raises lipophilicity from piracetam's strongly hydrophilic profile into the range that crosses membranes readily, which is also why the practical advice to take it with food is correct.

An identity warning that matters for anyone buying powder. Sunifiram is also C14H18N2O2 with a molecular weight of 246.30 National Center for Biotechnology Information 2026. The two are constitutional isomers with entirely different scaffolds and a roughly thirtyfold difference in dose. A certificate of analysis showing mass 246.3 is consistent with either. Only NMR or chromatography against a reference standard distinguishes them.

The mechanism was worked out by patch clamp, and it is more specific than “broad receptor profile” suggests. Moriguchi 2007 used whole-cell recording and immunoblotting in cultured rat cortical and hippocampal neurons and established four things. First, nefiracetam's potentiation of NMDA currents was blocked by the protein kinase C inhibitor chelerythrine and not by the protein kinase A inhibitor H89 — and PKA activity was unaffected. Second, nefiracetam raised PKC-alpha activity with a bell-shaped dose-response peaking at 10 nM, increasing phosphorylation of PKC substrates and of the NMDA receptor itself. Third, current-voltage analysis showed that at 10 nM it largely eliminated the voltage-dependent magnesium block of the NMDA receptor, and that this too was PKC-dependent. Fourth, and the authors state it explicitly: it potentiates not by acting as a partial agonist but by interacting with PKC, allosterically enhancing glycine binding and attenuating the Mg²⁺ block.

Removing the magnesium block is a bigger deal than it sounds. The Mg²⁺ plug is what makes the NMDA receptor a coincidence detector: it only conducts when glutamate is bound and the membrane is already depolarized. A drug that attenuates that block lowers the threshold for NMDA current at resting potential, which is a plausible route to enhanced synaptic plasticity — and, at higher exposure, the same property that makes excessive NMDA signaling a problem.

The nicotinic arm is real and was characterized first. Moriguchi 2007's own introduction records the group's earlier finding that nefiracetam potentiates both nicotinic acetylcholine receptors and NMDA receptors, and modulates the glycine binding site. Two receptor systems, one kinase pathway. The mood effect people report is usually attributed to a GABAergic action; the published signaling work points instead at PKC-alpha convergence on nicotinic and NMDA receptors, which is a different and better-evidenced story.

Cell, rodent, human — and where it stops

Step one, neurons in culture. Rat cortical and hippocampal neurons, whole-cell patch clamp, effect concentration 10 nM, bell-shaped Moriguchi 2007. Hold that number.

Step two, human pharmacokinetics, which are unusually complete for a compound in this catalog. Fujimaki 1992 dosed healthy Japanese male volunteers with 10 to 200 mg orally as single doses and 200 mg three times daily for seven days. Kinetics were linear, Cmax arrived within 2 hours, the half-life was 3 to 5 hours, apparent clearance was 94.4–140.3 mL/min, less than 10% of the dose appeared unchanged in urine at 24 hours and less than 0.1% in feces, food delayed absorption without changing the overall pharmacokinetics, and there was no clinically significant accumulation on repeat dosing.

Now put step one and step two beside each other, because this is the translational problem of the whole compound. Fujimaki 1993 measured a peak plasma concentration of 16.3 ± 0.9 nmol/mL after a single 200 mg dose. That is 16.3 µM — roughly 1,600 times the 10 nM at which PKC-alpha activation peaked in culture, on a dose-response curve that declines above its peak Moriguchi 2007. Protein binding and brain-to-plasma partitioning will narrow that gap and neither has been published for this molecule, so the true multiple is unknown. But three orders of magnitude is not closed by protein binding alone, and the possibility that clinical doses sit far past the optimum has never been tested.

Step three, the metabolites, which outlive the parent. Fujimaki 1993 found parent half-life 3.9 hours with tmax 1.6 hours, and three metabolites with half-lives of 7.8 to 21.9 hours and tmax of 4.1 to 9.6 hours, formed by stereoselective hydroxylation of the pyrrolidine ring — one metabolite a racemate, another predominantly the (−) enantiomer in human urine. Total drug-related material in urine reached 43.4% of the dose. So the compound a person carries through the afternoon is mostly metabolite, and none of the metabolites has ever been tested for activity.

Step four, the human trials — one missed and one hit. Robinson 2008 enrolled 159 patients within three months of a stroke with major depression into a double-blind trial of nefiracetam against placebo. The primary analysis failed: repeated-measures analysis of covariance showed no significant time-by-treatment interaction, and response rates exceeded 70% with remission above 40% in both arms, which is what a large placebo response looks like. Only the top quintile of Hamilton Depression scores showed a significant effect, at 900 mg versus 600 mg or placebo. The authors' own conclusion: not an effective treatment for post-stroke depression.

Robinson 2009 then looked at apathy inside the same program. Of 137 stroke patients with major depression, 70 (51.1%) also met criteria for apathy and were randomized to placebo, 600 mg or 900 mg daily for at least four weeks. Here the repeated-measures analysis of Apathy Scale scores did show a significant time-by-treatment interaction, with 900 mg significantly better than 600 mg or placebo. That is a positive result on a prespecified scale in a subgroup, and the authors' closing recommendation was to test apathy without depression — a study that was never run.

The obstacle. The positive finding is a subgroup of 70 inside a trial whose primary endpoint failed, in stroke patients, at a dose (900 mg) three to nine times what this site's card lists. The population is not healthy adults, the endpoint is apathy rather than cognition, and the only cognitive-enhancement claim rests on the cell-culture work at a concentration a thousandfold below what the plasma actually contains.

Nefiracetam pharmacokinetics — how much of it actually gets in

This is one of the few research compounds in the vault with a proper published human pharmacokinetic study, and it is worth using.

Route and absorption. Oral. Cmax within 2 hours, linear kinetics from 10 to 200 mg, apparent clearance 94.4–140.3 mL/min Fujimaki 1992. Food delays absorption without significantly changing the overall exposure — so the practical instruction to take it with a meal is about tolerability and smoothing the peak, not about improving oral bioavailability. With XLogP 1.4 and TPSA 49.4 National Center for Biotechnology Information 2026, passive absorption and brain penetration are both expected to be good.

What degrades it. Hepatic oxidation. Fujimaki 1993 identified stereoselective hydroxylation of the pyrrolidine ring as the principal route, producing three metabolites; the cytochrome responsible has not been named in the human literature. Renal excretion of unchanged drug is minor — under 10% at 24 hours — and fecal excretion is negligible at under 0.1% Fujimaki 1992, with total drug-related material in urine reaching 43.4% of the dose Fujimaki 1993.

Two half-lives, and the second one is the one nobody accounts for. Parent half-life is 3.9 hours Fujimaki 1993, consistent with the 3–5 hours from the single-dose study Fujimaki 1992. The metabolites run 7.8 to 21.9 hours and peak 4.1 to 9.6 hours after the dose. So a twice-daily schedule keeps parent concentrations pulsatile and metabolite concentrations relatively flat, and since no metabolite has been assayed for activity, nobody knows which of those two curves the effect follows. Despite that, seven days of 200 mg three times daily produced no clinically significant accumulation Fujimaki 1992 — which is a reassuring finding and an unusual one for a compound with 22-hour metabolites.

The concentration to keep in mind. A 200 mg dose produced a peak of 16.3 µM Fujimaki 1993 against an in-vitro optimum of 10 nM Moriguchi 2007. There is no injectable formulation and no reason to want one: absorption is not the limitation here, and the open question is whether the oral dose is already far above the concentration the mechanism was characterized at.

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

Four predictions. The first comes straight out of the dog toxicology and is the most valuable unmeasured signal attached to any compound in this cohort.

1. The dog testicular finding produces an exact human signature, and it is three cheap tests. Shimomura 2004 gave beagle dogs 180 or 300 mg/kg/day and found a specific endocrine pattern: testicular testosterone fell within 4 hours of a single 300 mg/kg dose while progesterone did not change; serum testosterone fell after single, 1-week and 2-week dosing; serum estradiol rose from week 1 through week 4; and LH, FSH and inhibin B did not move at all throughout. The authors' interpretation is impaired conversion of progesterone to testosterone in the Leydig cell — a block at a specific steroidogenic step, below the pituitary. So the human test is Total Testosterone, Estradiol, Sensitive (LC/MS-MS) and LH & FSH at baseline and at 8 weeks. Testosterone down and estradiol up with LH and FSH unchanged is the dog signature. If the axis were being suppressed centrally instead, LH would fall; if nothing is happening, all four are flat. This has never been drawn in a human on nefiracetam, and it is three lines on a standard requisition.

2. The prediction that cuts against the compound: more should not be better, and might be worse. Moriguchi 2007 found a bell-shaped PKC-alpha response peaking at 10 nM, while a 200 mg dose reaches 16.3 µM in plasma Fujimaki 1993. If the inverted-U survives in vivo, a 100 mg dose should perform at least as well as 300 mg. Test it within one person: 100 mg and 300 mg on separate days, randomized, with Trail Making A and B at 2 and 6 hours. A monotonic dose-response falsifies the concern; a flat or inverted one confirms that the standard doses are chosen by habit rather than by pharmacology.

3. The effect should be on drive, not on memory. The one positive human result is on the Apathy Scale, at 900 mg Robinson 2009, in a trial whose depression endpoint failed Robinson 2008. So the honest expectation is a motivational effect rather than a mnemonic one, and the right instrument is an apathy or motivation scale plus a simple behavioral count — tasks initiated per day — not a memory test. Measuring the wrong thing is why most self-experiments with this compound produce no signal.

4. Safety chemistry belongs in the same draw. A Comprehensive Metabolic Panel (CMP) at baseline and 8 weeks. The dog program also found bladder and kidney lesions at high doses, and there is no human chronic-dosing safety database at all. Expected result: no change. Any change is new information about a compound nobody has monitored.

What nobody has tested yet

Nobody has measured testosterone in a human taking nefiracetam. This is the standout gap. Shimomura 2004 produced a mechanistically precise, dose-dependent, time-resolved endocrine lesion in dogs — testicular testosterone down at 4 hours, serum estradiol up by week 1, gonadotropins untouched, seminiferous atrophy with multinucleated giant cells by week 4 — and it produced an exact set of blood markers to look for. In the thirty years since nefiracetam entered human trials, no published study reports a testosterone value from a person taking it. Twenty men, an eight-week course, and four tubes of blood would close a gap that currently forces every discussion of this compound into speculation.

Nobody has tested the metabolites. Three hydroxylated metabolites with half-lives of 7.8–21.9 hours circulate for far longer than the parent Fujimaki 1993, and not one has been assayed against the NMDA or nicotinic receptor. Synthesizing them and running them through Moriguchi 2007's patch clamp preparation would determine whether the drug people experience in the evening is the molecule they swallowed in the morning.

Nobody ran the trial the apathy paper asked for. Robinson 2009 closes by recommending a study of apathy without depression — the cleaner population, without the enormous placebo response that sank the depression endpoint. Sixty patients with post-stroke apathy and no major depression, randomized to 900 mg, 600 mg or placebo for twelve weeks with the Apathy Scale as the primary outcome, is a straightforward and inexpensive trial. It has not been done, and the compound's development stopped instead.

And nobody has resolved the thousand-fold concentration gap. A study measuring both plasma and cerebrospinal fluid nefiracetam alongside a cognitive endpoint, across 100, 300 and 900 mg, would establish whether brain concentrations approach the 10 nM optimum or blow past it — and would either rescue the mechanistic account or force it to be rewritten.

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

The reason this never became a drug is a testicular finding in dogs, and it deserves the full detail rather than a one-line warning. Shimomura 2004 dosed beagle dogs at 180 or 300 mg/kg/day. After 4 weeks — but not after 1 week — histopathology showed moderate seminiferous atrophy at 180 mg/kg and severe atrophy with multinucleated giant cell formation at 300 mg/kg. Semen showed decreased motility and increased malformed sperm, also first at 4 weeks. The endocrine sequence ran ahead of the histology: testicular testosterone fell within 4 hours of a single high dose with progesterone unchanged, serum testosterone fell from the first dose onward, serum estradiol rose from week 1 to week 4, and LH, FSH and inhibin B never moved. The authors identify serum testosterone as the sensitive early predictor and propose a block in the conversion of progesterone to testosterone in the Leydig cell.

Now the margin, stated fairly in both directions. The dog doses were 180–300 mg/kg/day. A person taking 300 mg a day at 70 kg is taking about 4.3 mg/kg — roughly 40 to 70 times lower on a body-weight basis, and further still on a surface-area basis. That is a substantial margin and it is the reason the compound reached phase 2 in people at all. What the margin does not do is make the finding irrelevant, for two reasons: the mechanism is a steroidogenic block rather than a nonspecific toxicity, and mechanisms of that kind often have shallow dose-response relationships at the biochemical level even when frank histology needs a high dose. And nobody has ever measured the relevant hormones in a person, so the margin is a calculation rather than an observation.

The dog program found more than the testes. The same toxicology series included examination of bladder and kidney lesions in dogs. Those findings, like the testicular ones, were at high doses in a species with its own sensitivities — but they are the reason a plain metabolic panel is the right routine monitoring here rather than nothing.

The efficacy side of the risk-benefit is thinner than most summaries admit. The registered trial in post-stroke depression failed its primary analysis with response rates above 70% in both arms Robinson 2008; the positive apathy result is a prespecified analysis in 70 patients at 900 mg/day Robinson 2009, which is three times the top of the dose range quoted for research use. So the evidence is: one failed trial, one positive subgroup at a much higher dose, a dog testicular lesion, and no human endocrine monitoring at all.

Two practical points that follow from the pharmacokinetics. Seven days of 200 mg three times daily produced no clinically significant accumulation Fujimaki 1992, so the drug does not creep up on a daily schedule — a genuine point in its favor. But metabolites with half-lives up to 21.9 hours Fujimaki 1993 do persist across the day and have never been characterized for activity or toxicity, which is where the honest uncertainty sits.

Sources read for this page

• Moriguchi S, Shioda N, Maejima H. Nefiracetam potentiates N-methyl-D-aspartate (NMDA) receptor function via protein kinase C activation and reduces magnesium block of NMDA receptor.. Mol Pharmacol 2007 · PMID 17095583
• Fujimaki Y, Sudo K, Hakusui H. Single- and multiple-dose pharmacokinetics of nefiracetam, a new nootropic agent, in healthy volunteers.. J Pharm Pharmacol 1992 · PMID 1360528
• Fujimaki Y, Sudo K, Hakusui H. Pharmacokinetics of nefiracetam and three metabolites in humans and stereoselective hydroxylation of its pyrrolidine ring.. Xenobiotica 1993 · PMID 8484264
• Robinson RG, Jorge RE, Clarence-Smith K. Double-blind randomized treatment of poststroke depression using nefiracetam.. J Neuropsychiatry Clin Neurosci 2008 · PMID 18451188
• Robinson RG, Jorge RE, Clarence-Smith K. Double-blind treatment of apathy in patients with poststroke depression using nefiracetam.. J Neuropsychiatry Clin Neurosci 2009 · PMID 19622685
• Shimomura K, Shimada M, Hagiwara M. Testicular toxicity induced in dogs by nefiracetam, a neutrotransmission enhancer.. Reprod Toxicol 2004 · PMID 15082078
• National Center for Biotechnology Information. PubChem Compound Summary for CID 71157, Nefiracetam.. PubChem, retrieved 2026
• National Center for Biotechnology Information. PubChem Compound Summary for CID 4223812, Sunifiram.. PubChem, retrieved 2026

Nefiracetam — safety, predicted from mechanism

Nefiracetam — interference & stacking

Bloodwork to run alongside Nefiracetam

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

Nefiracetam — frequently asked questions

What Nefiracetam is used for

Nefiracetam appears under 1 goal in the goal router.

Related Cognitive & Mood compounds

← Explore the full Protocol Vault

↑ Back to on this page