Mexidol, bemethyl and phenylpiracetam

01Key questions

02What these compounds are

Swipe table →

03How to read this literature

Almost every positive clinical paper on Mexidol and phenylpiracetam appears in one Russian journal, the Zhurnal Nevrologii i Psikhiatrii imeni S. S. Korsakova. It is a real, indexed journal. But when nearly all the good news for a drug comes from one journal and one country, and the sponsor is rarely named in the abstract, independent replication is missing. That is not a reason to dismiss the work. Western regulators ignore much of it for a different reason: nobody with a patent to protect will pay for a Western trial of an old Russian molecule (see off-patent compounds and missing trials). The fair reading is the one used on Russian pharmacology: check the chemical identity, the mechanism at real-world doses, the human evidence beyond custom, and the governance (sponsor, doping status, dependence).

04Mexidol: two stroke trials

EPICA (2017)

A randomized, double-blind, placebo-controlled, multicenter trial in 150 adults aged 40 to 79 with hemispheric ischemic stroke (PMID 28665371). Patients received Mexidol by intravenous infusion for 10 days and then tablets for 8 weeks, or a matching placebo, on top of standard care. At the end of treatment the Mexidol group had a lower mean disability score on the modified Rankin Scale (mRS, p=0.04), more patients with little or no disability (mRS 0 to 2, p=0.039) and a lower stroke-severity score on the NIHSS (p=0.035). Side effects were no more common than on placebo. The abstract gives p-values but not the scores themselves or the size of the difference, and it does not name a sponsor.

A 2020 re-analysis of the same patients split them by age (PMID 33016677). It reported the clearest difference from placebo in the oldest group, aged 76 to 90 (p<0.001). The English and Russian versions of that abstract give different limits for the middle age band (60 to 65 versus 60 to 75), so that subgroup result cannot be pinned down from the abstract.

MIR (2025)

The larger trial: 304 patients at 17 centers in Russia, Kazakhstan and Uzbekistan, randomized and double-blind, adults aged 40 to 75 with a first hemispheric ischemic stroke treated within 48 hours and without clot-removal or clot-dissolving therapy (PMID 40898634; PMID 41456191; NCT06437626). Mexidol was given intravenously for 10 days and then as tablets for 60 days, again on top of standard care.

The ClinicalTrials.gov record, which has posted results, names Pharmasoft as the lead industry sponsor. The journal abstracts do not. Here is what the registry shows at about day 71:

• Disability (primary outcome): mean mRS improved by 2.45 points on Mexidol (n=141) and 2.01 on placebo (n=138). The difference is 0.44 points on a 0 to 6 scale. The companion paper reports p=0.003.
• Little or no disability (mRS 0 to 1): 85 of 141 on Mexidol versus 57 of 138 on placebo (60% versus 41%).
• Severe disability (mRS above 3): 22 of 141 versus 39 of 138.
• Stroke severity (NIHSS): the median improvement was 7 points in both groups. The primary paper still reports a significant NIHSS difference (p<0.001); the registry's identical medians do not show one.
• Mobility and cognition: median gains of 10 versus 9 points on the Rivermead Mobility Index and 4 versus 3 points on the MoCA cognitive test. The authors call the MoCA result a trend, not a significant finding.
• Safety: adverse events in 23% of each group. Two deaths on Mexidol and one on placebo; six versus four patients with a serious adverse event. These numbers are too small to compare.

Two smaller discrepancies remain between the paper and the registry: the primary paper describes its main contrast with medians while the registry reports means, and the upper confidence limit for the Mexidol group is given as −2.32 in the paper and −2.23 in the registry.

Taken together, Mexidol has two positive placebo-controlled trials in acute stroke, which is more than most compounds on this site. Both come from the same research network and journal, the larger one is industry-funded, and the main effect is a modest shift on the disability scale. A trial run by an independent group, ideally outside Russia, would settle it. Neither trial tested Mexidol as a cognitive enhancer in healthy people, which is how it is usually sold online.

What is not a Mexidol trial

Two 2020 papers in the same journal (PMID 32678558, PMID 33449535) are reviews, not new patient data. The second one says early treatment within six hours works best; neither trial tested that, and MIR enrolled patients up to 48 hours after onset. A 2022 review (PMID 35992374) covers patents, iron binding in the lab and antioxidant theory. The Italian "EPICA" atrial-fibrillation studies are unrelated and only share the acronym. ClinicalTrials.gov lists other completed Mexidol studies (chronic cerebral ischemia, childhood ADHD, vascular cognitive impairment, glaucoma), but only MIR has posted results.

05Bemethyl: almost no human data

Bemethyl was developed in the Soviet Union as an "actoprotector", a drug meant to keep soldiers and workers performing under heavy physical load. Its reputation rests on that history, not on published trials. ClinicalTrials.gov has no registered trial under any of its names, and PubMed has no human cognition trial.

The human records that do exist are thin:

• A pharmacokinetic study in healthy volunteers after a single 250 mg capsule found peak blood levels about one hour after the dose and very uneven levels between people (PMID 21870773). The abstract gives no number of participants or half-life.
• A 1997 study in men doing physical work in insulated protective suits reported that one dose of bemethyl or phenibut raised heat tolerance, with phenibut the stronger of the two (PMID 9162292). No sample size is given.
• A study after brain injury tested bemethyl only in combination with pyrazidol, so it says nothing about bemethyl alone (PMID 16404942).

A 2021 rat study mapped nine breakdown products in urine, mainly for drug testing (PMID 34445727). It notes that WADA has had bemethyl on its Monitoring Program since 2018. That program tracks substances for possible future bans; it is not a ban.

06Phenylpiracetam (fonturacetam)

Human studies

• Epilepsy, add-on (placebo-controlled): 90 adults with focal epilepsy took phenylpiracetam or placebo alongside their usual medicines (PMID 25591651). The authors report fewer seizures and better cognition in patients without epileptiform EEG changes, but the abstract gives no arm sizes, seizure counts or p-values. In 40% of patients it did not reduce the side effects of the epilepsy drugs.
• After brain lesions: 99 patients treated for a month, with no control group (PMID 16447562). It describes improvement over time, which cannot be separated from natural recovery.
• Heart patients with anxiety or low mood: an open comparison of 35 patients on phenylpiracetam and 35 on another Russian drug, pantogam active, for 12 weeks, with no placebo (PMID 25403298). Pantogam active did better on several measures, and the calming effect of phenylpiracetam faded after four to eight weeks.
• Fatigue (asthenia), 2025 meta-analysis: 11 studies, 549 patients, all at the same daily dose (PMID 40047835). Fatigue scores (MFI-20) fell by an average of 16.3 points after a month (95% CI 8.85 to 23.85), and 5.5% had side effects described as transient. This is a before-and-after change in treated patients. The abstract does not say whether any included study had a placebo group, so the result cannot be read as an effect of the drug. The paper's objective names the Actitropil brand.

Phenylpiracetam has no trial in healthy people, the group that buys it as a nootropic.

Animal work

In mice, the R-form increased movement and improved memory in a passive-avoidance test, and the S-form did not improve memory (PMID 21689376). Rat and mouse studies compare it with piracetam after reduced blood flow to the brain (PMID 17523446) and test the R-form against inflammation (PMID 32279140). Animal doses do not translate to human doses, and the claim that the R-form is "the active one" in people rests on this mouse work alone.

Banned in sport

The 2026 WADA Prohibited List names fonturacetam (4-phenylpiracetam, carphedon) as a non-specified stimulant, alongside modafinil. It is prohibited in competition. Whether the asthenia studies are positive is a separate question from its doping status.

07Cortexin and Pinealon

These two are often sold alongside the drugs above. Cortexin is a peptide extract from animal brain cortex, a mixture rather than a single molecule, so a result from one batch does not automatically carry over to another; its stroke studies are covered on Russian pharmacology and Cerebrolysin. Pinealon (Glu-Asp-Arg) is a Khavinson bioregulator peptide with cell, rat and computer-model studies but no human trial under that name; see Khavinson peptide bioregulation.

08Claims versus evidence

Swipe table →

09Practical points

• The stroke evidence for Mexidol used the pharmaceutical product, given intravenously and then by mouth under medical care. Emoxypine powders sold online are not the same preparation.
• Phenylpiracetam is prohibited in competition. Athletes who are tested should treat it as a doping substance.
• Bemethyl is not bromantane, and neither has trial data for cognition.
• None of these drugs has FDA approval. Products sold online as research chemicals are not the pharmaceutical preparations used in the trials.

What studies and practitioners report

Amounts as reported in the cited sources: trial doses with the population studied, and practitioner-reported ranges marked as such. Not medical advice.

• Trial 500 mg/day IV for 10 days, then 125 mg three times daily by mouth for 8 weeks · Mexidol, EPICA stroke trial adults with ischemic stroke, n=150 PMID 28665371
• Trial 500 mg twice daily IV for 10 days, then 250 mg three times daily by mouth for 60 days · Mexidol, MIR stroke trial adults with first ischemic stroke, n=304 PMID 40898634
• Trial 100 or 200 mg · phenylpiracetam, add-on epilepsy trial adults with focal epilepsy, n=90 PMID 25591651
• Trial 200 mg/day for one month · fonturacetam, asthenia meta-analysis (all included studies) patients with asthenia, n=549 PMID 40047835
• Trial 250 mg, single oral dose · bemethyl (Metaprot), pharmacokinetic study healthy volunteers PMID 21870773

11Connections

12Sources

1. PubMed 35992374PubMed 35992374
2. PubMed 21870773PubMed 21870773
3. PubMed 21689376PubMed 21689376
4. PubMed 28665371PubMed 28665371
5. PubMed 33016677PubMed 33016677
6. PubMed 40898634PubMed 40898634
7. PubMed 41456191PubMed 41456191
8. NCT06437626NCT06437626
9. PubMed 32678558PubMed 32678558
10. PubMed 33449535PubMed 33449535
11. PubMed 9162292PubMed 9162292
12. PubMed 16404942PubMed 16404942
13. PubMed 34445727PubMed 34445727
14. PubMed 25591651PubMed 25591651
15. PubMed 16447562PubMed 16447562
16. PubMed 25403298PubMed 25403298
17. PubMed 40047835PubMed 40047835
18. PubMed 17523446PubMed 17523446
19. PubMed 32279140PubMed 32279140

13Entry history

Heretic informs and never prescribes. Where amounts appear, they describe what published studies used. Talk to a qualified clinician before acting on anything here.