Compound guide / acetylcholine
Seven dosed nights at 8mg on a wake-back-to-bed schedule, every other night over fourteen days. Three lucid dreams, the clearest signal in the matrix so far, and a set of limits that matter as much as the result.
Galantamine is an acetylcholinesterase inhibitor, originally isolated from snowdrop (Galanthus nivalis) and used clinically in Alzheimer's treatment. It slows the breakdown of acetylcholine, the neurotransmitter most closely tied to REM sleep and to dream vividness.
That mechanism is why it appears in lucid dreaming discussions at all. It is not a sedative and it is not a dream herb in the ethnobotanical sense. It is a drug with a known pharmacological target that happens to sit near the chemistry of REM.
Seven dosed nights across fourteen days, with a washout day between each. Eight milligrams, taken on waking roughly four and a half hours into sleep, then back to bed. Fixed dose, fixed timing, no stacking, no behavioural techniques layered on top, no dream journaling introduced as a second variable.
The point of that rigidity is that dreams vary enormously night to night. A single good night proves close to nothing. Only a run at fixed conditions gives you anything to compare against a baseline.
Three lucid dreams across the run, clustered on nights five through seven rather than spread evenly. Recall ran high throughout but fragmented, with more remembered material and less narrative continuity holding it together. The back half of the night felt lighter, which is consistent with the mechanism rather than evidence for it.
Against a lifetime count of ten lucid dreams, three inside one seven night run is the reason galantamine is the only compound in the matrix carrying a high impact rating.
The seven night acute phase of this run is documented on video: the dose, the timing, and the nights that produced lucidity, in about six minutes.
It did not produce lucidity on command. Lucidity showed up on a minority of nights even during an active run, and the clustering pattern suggests something building and then fading rather than a per-night switch.
It also did not hold up as something to repeat indefinitely. Tolerance builds fast enough that nightly dosing is the wrong approach, which puts a ceiling on how much any single person can learn from it.
And this remains one person, unblinded, aware of what was being tested, with no control arm. Expectancy alone can produce lucid dreams in someone paying close attention to their sleep. The result is a signal, not a finding.
Twenty-two compounds have been through the matrix. Five coincided with at least one lucid dream. Eight produced no measurable change at all. Galantamine is the only one that produced a repeatable cluster rather than an isolated event.
The two other acetylcholine-adjacent compounds tested so far, huperzine A at 200 and 400mcg and alpha-GPC at 600mg, both on the same wake-back-to-bed timing, produced nothing comparable. That is worth sitting with. If the mechanism were the whole story, the family should behave more alike than it does.
Alpha-GPC is a choline source, so this was the galantamine and choline pairing people ask about. Pairing galantamine with alpha-GPC did not help. The galantamine and Alpha-GPC stack ran twice at 8mg and 600mg on the same timing, seven dosed nights each, and produced zero lucid dreams. Recall improved in the second run, but lucidity per night fell below galantamine alone.
Answered only from our own runs. None of this is dosing advice.
In our seven night run, dosed every other night over fourteen days, it coincided with three lucid dreams, clustered on nights five through seven. That is the strongest signal in our matrix so far. It is one person, unblinded, with no control arm, so treat it as a signal worth chasing rather than proof of cause.
Not in our testing. Alpha-GPC is a choline source, and pairing 600mg of it with 8mg of galantamine at the same wake-back-to-bed timing produced zero lucid dreams across two seven night runs. Galantamine alone produced three. Alpha-GPC on its own was a null result. Adding choline is common advice, and our data gives no support for it.
It was for us. Huperzine A at 200 and 400mcg, on the same wake-back-to-bed timing, produced no lucid dreams, against three for galantamine. Both slow the breakdown of acetylcholine, so the gap is not explained by mechanism alone. It may come down to dose, how long each compound stays active, or chance in a single subject.
REM density climbs across the back half of the night. Dosing at the wake-back-to-bed point, roughly four and a half hours into sleep, places the compound near that window rather than during early slow wave sleep, when it would have less to act on.
Eight milligrams, taken on a wake-back-to-bed schedule. We did not test other doses, so we cannot say whether four or sixteen would behave differently.
We would not. Tolerance builds fast enough that nightly dosing looks like the wrong move, and the nights that produced lucidity were a minority even during an active run.
Galantamine is a prescription drug in most jurisdictions and a cholinergic one. Nausea is common. It interacts with other cholinergic and anticholinergic medications, and it is not appropriate for people with certain cardiac conduction problems, asthma, or ulcers. Nightcone documents self-experimentation for research interest. Nothing here is a recommendation to take anything.
The open questions are timing and spacing rather than dose. Whether the cluster on nights five through seven reflects accumulation or coincidence, whether a rest day pattern preserves the effect longer than consecutive nights, and whether the wake-back-to-bed interval itself is doing more work than the compound. Those are conditions, not chemistry, and conditions are cheaper to test.
Results from controlled oneirogen experiments. What impacted dreaming, what didn't, and topical news. New tests each cycle.