Semax: A Seven-Amino-Acid Peptide With a Stroke Trial Behind It

ACTH stands for adrenocorticotropic hormone. Your pituitary makes it. Its best-known job is telling the adrenal glands to release cortisol.

Decades ago, researchers noticed something else. Short fragments of ACTH seemed to affect learning and attention without carrying the hormone's other effects. Semax was built on that idea.¹

It copies ACTH(4-10), the fourth through tenth amino acids of the hormone, with a modification that helps it last longer.¹

The whole sequence is short enough to print: Met-Glu-His-Phe-Pro-Gly-Pro.² The Pro-Gly-Pro tail at the end is the added part. It is what separates semax from the plain hormone fragment.

This is the largest clinical dataset in the published record. It is worth reading precisely, not in summary.

One hundred and ten patients were examined after ischaemic stroke. That was 43 men and 67 women, mean age 58.³

They were split by when rehabilitation started. An early group began at about 89 days. A late group began at about 214 days.³

Each group was then split again, into those who received semax and those who did not.³

Researchers measured three things. Plasma BDNF. Motor performance, on the British Medical Research Council scale. And the Barthel index, which scores how independently a person manages daily life.³

The reported result: semax raised plasma BDNF levels, and it did so regardless of whether rehabilitation had started early or late.³

One limitation is structural. The paper is in Russian, in a Russian neurology journal. That does not make it wrong. It does mean far fewer outside readers have checked it than would have checked an English trial of the same size.

The most thorough mechanism work went genome-wide. Researchers induced focal ischaemia in rats. Then they compared gene expression in treated and untreated animals, at 3 and 24 hours.²

The result is not the one a simple neuroprotection story predicts. Semax mainly raised the expression of immune-system genes. At 3 hours it affected genes governing the activity of immune cells.²

Genes tied to the vascular system showed up too.² The picture is of a compound acting on the brain's whole response to injury, not on neurons alone.

Other animal work looks past stroke. One study asked whether semax reduced behavioural and neurochemical changes in rats exposed to a stressor early in life.¹ A 2025 review gathers the animal-model literature in one place.⁴

One study scanned semax in healthy volunteers. Fifty-two participants had resting-state fMRI before, then at 5 and at 20 minutes.⁵

The regions were chosen in advance: the amygdala, central to anxiety, and the dorsolateral prefrontal cortex, which handles executive function and working memory.⁵

Differences showed up between the right amygdala and parts of the right temporal cortex. The authors present these as the first described effects of semax and selank on that link.⁵

It is a mechanism finding in healthy people. It is not an outcome in patients, and the paper does not claim otherwise.

• It is not approved in the United States. Its regulatory history is Russian, and material sold here is research material.
• Claims about focus in healthy people are not what the clinical work tested. The stroke study enrolled patients recovering from a stroke.³ The scan study measured connectivity, not performance.⁵
• Raising BDNF is a mechanism, not a result. The study measured it as a marker alongside motor and daily-function scores.³
• Most mechanistic work is in rats. Genome-wide expression after experimental stroke is a rat model of one condition.²