Seltorexant

most data

Also seen as: JNJ-42847922, MIN-202

The Dose Guy’s verdict

Aug 29, 2026
3 sources

Seltorexant looks genuinely good in trials, matching or beating zolpidem on sleep and showing a promising signal for depression on top of it. But it is not approved, and until it is, this is a compound you watch, not one you run. If you want the orexin-blocking approach to sleep right now, lemborexant is the approved version that does the same job. Seltorexant’s edge, if it holds, is the depression angle: fixing sleep and mood from the same receptor.

Would I take it?Not yet. Run lemborexant instead.

What it is

A selective orexin-2 receptor antagonist, meaning it blocks only one of the two orexin receptors instead of both. Where lemborexant hits OX1R and OX2R, seltorexant leaves OX1R alone, which in theory gives a more targeted sleep signal with less daytime spillover. The development program is unusual: it is being pushed simultaneously for insomnia and as an adjunct for major depressive disorder, because bad sleep and depression share orexin wiring, and a drug that resets both could be genuinely useful.

It is still in late-stage trials. The insomnia and depression data is encouraging, but “encouraging phase 2” is a long way from “available and proven.” The story here is one to follow, not one to act on.

What the research shows

Phase 1b and 2 trials across insomnia and depression, all positive, none yet definitive:

In a 14-day insomnia trial, 20 mg cut sleep-onset latency to 51% of placebo on night 1, beat zolpidem by 29%, and maintained both gains through night 13.JAMA Psychiatry 2025 · phase 2 · n=364

As an add-on in MDD, 20 mg improved MADRS depression scores by 4.5 points over placebo at week 3, p=0.003, with a larger effect at week 6 in patients who also had significant insomnia, 4.9 versus 0.7 points by baseline Insomnia Severity Index.IJNP 2021 · phase 2b · adaptive

As MDD monotherapy, 20 mg improved HDRS-17 scores versus placebo at week 5, p=0.0049, and the benefit held after sleep items were stripped out.Mol Psychiatry 2025 · phase 1b · n=128

Realistic protocols

Seltorexant is not approved. There is no defensible self-directed protocol to write. The insomnia trials tested 5, 10, 20, and 40 mg nightly, with 20 mg showing the strongest signal, and the adjunctive depression trials used 20 mg. As with all orexin antagonists, alcohol stacks on the sedation. If you want orexin blockade for sleep today, lemborexant is the compound to run. If you source seltorexant from a research-chemical vendor anyway, start at the lowest tested dose, skip alcohol on treatment nights, and log your sleep and mood honestly.

Side effects

Trial-reported and mild so far, though long-term data is limited:

commonSomnolence, headache, and nausea, the usual orexin-antagonist profile.

monitorAsymptomatic ECG changes caused a handful of trial discontinuations. The clinical meaning is unclear, but worth watching if it reaches market.

unknownNo long-term safety data beyond 24 weeks of trial exposure. The class is young, and this member has less human mileage than lemborexant.

Where to buy

No live listings from tracked vendors right now.