Morning Glory

The danger of this seed is not its potency, it is its irregularity and the act it prompts: the person who feels nothing takes another dose. Its reputation as a natural LSD does not hold. The only work to have measured ergine at receptors concludes that it should not be regarded as an LSD-type psychedelic drug, and the only controlled human experiment, at 5 mg, produced nothing but drowsiness. The garden morning glory, Ipomoea tricolor, is an ornamental seed sold in garden centres under the cultivars Heavenly Blue, Pearly Gates and Flying Saucers, and its ergolines are not even made by the plant: they come from a symbiotic fungus, and only about one seed in three carries it. That is where the irregularity begins.

What the seed contains, and who makes it

The ergolines of the morning glories are produced by fungi of the genus Periglandula, epibionts of the Convolvulaceae, which carry the gene for the determining step of the biosynthesis. The symbiont specific to I. tricolor was isolated in 2025 under the name Periglandula clandestina, the epithet referring to its hidden character: it is not visible on the leaves, only in the empty seed coats. Treatment with the fungicide tebuconazole abolishes ergoline accumulation, which demonstrates the fungal origin, and the same work finds the identical symbiont in the three cultivars. Three ergolines are detected there: lysergic acid alpha-hydroxyethylamide, known as LAH or LSH, ergine, in other words LSA, and ergonovine, in other words ergometrine.

Which one dominates in the seed remains open, and it must not be settled. LAH accounts for 58 to 77% of the ergolines in that work, but the measurement is on vegetative tissue from six-week-old seedlings in hydroponics, the seeds were not assayed, and the quantification there is relative to an ergonovine standard. Other publications say the opposite for the seed. An analytical trap is added to this: LAH rearranges into ergine in an over-aggressive ionisation source and under microwave extraction, and ergine is moreover a hydrolysis product of LAH, sensitive to the age and the storage of the sample. Part of the published ergine figures is therefore probably converted LAH. Meanwhile, all the available pharmacology is on ergine, and no receptor characterisation of LAH exists. It is possible that this page, in speaking of LSA, names the wrong molecule for this particular plant.

What ergine does, and what nobody has measured

A single published work measures ergine at receptors. In pKi: 7.99 at 5-HT1A, 7.56 at 5-HT2, 7.21 at the alpha-2 adrenergic, 6.05 to 6.85 at the dopaminergic and alpha-1 subtypes. The authors state that LSA shows lower affinity than LSD at all the subtypes tested, and their conclusion is, word for word: it should not be considered an LSD-type psychedelic drug. The same article carried an in silico model whose main prediction was refuted by the measurement; its 5-HT1D, 5-HT1B, 5-HT6, 5-HT7 and D2 values remain purely predicted and must never be quoted as affinities.

There is no functional efficacy data for ergine at 5-HT2A: binding, not function. Partial agonist, weak agonist or antagonist, undetermined. Applying the 5-HT2A model of the classical hallucinogens to ergine is an extrapolation, not a result. Nor is there any system-level data in humans, no functional imaging, no EEG, no connectivity; the only published imaging connected with LSA is diagnostic, that of a posterior reversible encephalopathy syndrome. The entire literature amounts to very little: the most recent systematic review retained only 17 studies and 219 patients, without a single controlled trial or a single human dose-response curve.

The human experimental argument points the same way. In 1966, Isbell and Gorodetzky administered to human subjects either a crude extract of ololiuqui seeds or a mixture of synthetic alkaloids, at 5 mg: drowsiness, few other subjective effects, rarely perceptual distortion, neither fever nor marked mydriasis, no hallucination, whereas LSD-25 produced in the same setting nervousness, distortion, euphoria and hallucinations. Their conclusion is that ololiuqui and its alkaloids should be regarded chiefly as sedatives rather than as psychotomimetics. The reservations are heavy and inseparable: the article has not been read, the PubMed record carries no abstract, the one we have comes from a search engine and is corroborated only by the MeSH indexing of the record; the population consisted of former morphine addicts at the Addiction Research Center in Lexington; the sample size is unknown. It is the most important human source on this page and the least soundly held.

The only modern human study is on another plant, Argyreia nervosa, the Hawaiian baby woodrose, which has its own page, and it was halted. The same goes for very nearly the whole clinical corpus and the available blood measurements: they come from this neighbour, and each time a figure comes from it, that is said. Four volunteers, seeds by the oral route, serum LSA of 0.66 to 3.15 ng/mL about two hours after ingestion. At these concentrations: severe autonomic effects, nausea, weakness, fatigue, tremor, raised blood pressure, and a psychotic-type state that ended the study. The same team elsewhere describes the psychedelic activity observed as weak. That is the picture the sources describe, an autonomic component that dominates and a perceptual component weaker and more uncertain than the name LSA leads one to expect. Mydriasis, often cited, is documented only for Argyreia, and the closest source for Ipomoea explicitly says the opposite.

The dose is not a riddle, it is a lottery

Three independent estimates converge on about 4 to 23 µg of ergine per seed: a commercial lot at 0.011% LSA, a police seizure lot at 0.062%, and a derivation anchored on a sentence by authors who make the seed twelve times more concentrated than the green parts, that is about 295 µg/g of ergine and 19 µg/g of ergometrine. In other words, 250 to 400 seeds represent roughly 1 to 9 mg of ergine, that is to say Isbell's 5 mg dose. The folk doses do reach the experimental dose, and that dose does not produce hallucinations. There is no riddle to be solved there.

What remains is a variability quantified four times, and that is the real subject of this page.

• The weight of a seed varies from 15 to 65 mg. A factor of 4 before content is even mentioned: counting by number of seeds is structurally imprecise.
• A lot may contain nothing. Of five commercial lots analysed, only three contained ergolines, and all three carried the Heavenly Blue cultivar, although sold under two different species names.
• Two seeds from the same packet are not equivalent. Analysed one by one, they range from a value close to or below the detection limit, that is 1.0 µg/g, up to twice the mean. The authors name the risk themselves: overdose through accidental selection of more potent seeds, or taking a second dose when the first has acted less than the average.
• The mechanism is known. In a single lot of Heavenly Blue, 13 seed coats out of 37 carried fungal colonies and 14 seedlings out of 38 from the same lot were positive for ergolines. About one seed in three carries the symbiont, the rest is chemically inert. Reservation: seed coats observed after germination and seedlings assayed at four weeks in hydroponics, not an assay of dry seeds. It is the best figure available.
• The lottery is also played out between populations: in the positive species with at least three herbarium samples, ergolines are present in only 67% of the samples on average.

Two figures in circulation push in exactly the wrong direction. ICEERS gives for pure ergine an active dose of 1 to 2 mg by the oral route, attributed to Hofmann without a reference, which flatly contradicts Isbell's 5 mg and sets out from a supposed active threshold that is too low. The same page gives for I. violacea 5 to 10 grams, described as 6 to 13 seeds: it is the grams that are right and the seed count that is wrong, by a factor of about 12 to 50. A forum belief is added to these, according to which SSRIs strongly blunt the effect of LSA and force one to consume many more seeds: no primary source supports it. What is dangerous in these three statements is not what they assert, it is what they make people do, dose escalation on a material whose content already varies by a factor greater than two from one seed to the next. That is the central accident mode of this plant.

Duration and tolerance

Everything that follows concerns the oral route, whole or ground seeds, or an infusion, the only one that is documented.

ICEERS gives an onset between 40 and 90 minutes, a peak at 2 to 2.5 hours and a total duration of 5 to 8 hours, without a primary citation. The only compatible measured anchor is the serum peak at about two hours and urinary disappearance between 24 and 48 hours, on Argyreia. Treat the 5 to 8 hours as a harm reduction organisation's estimate, not as data.

In humans, ergine has neither a published plasma half-life nor an established enzymatic pathway: the 2-hydroxy-3-oxo metabolites of LSA and of ergonovine are identified in forensic toxicology, the enzyme that produces them is not. Nor has anything ever been measured of the metabolism of ergine by CYP3A4, which leaves open a real interaction question, ergotism through CYP3A4 inhibition being documented for ergotamine.

On tolerance, there is no data. The historical cross-tolerance programme of the Addiction Research Center covered psilocybin, mescaline, BOL-148, d-amphetamine and THC against LSD, never ergine. The common claim of cross-tolerance with LSD is a chemical-class extrapolation.

Harm reduction

The number one operational instruction fits in three words: do not redose. The three levels of lottery just seen stack up, and the act that follows from them is named by the analysts themselves. It holds regardless of any absence of felt effect, and it is precisely that absence which endangers it.

1. Suicidal ideation is the most repeated serious signal, and it comes before everything else. It recurs in four independent sources of different kinds: a case report of suicide after ingestion of morning glory seeds in 1964, a death by fall from a building with verbalisation of the intention to jump, two subjects in a corpus of user reports, a suicide after agitation and psychosis in the systematic review. None of these events is a dose toxicity: the risk of this plant is not a lethal dose, it is a behaviour. Induced psychosis is the second signal, most often self-resolving but with no established treatment, including in a person who has already used without incident; adolescents are over-represented in the case reports. Practical consequence: a sober third party, and no accessible window or balcony.

2. Pregnancy: contraindication, and it is the best supported point on this page. The seeds contain ergometrine, in other words ergonovine, the reference uterotonic, which is among the three ergolines detected in I. tricolor and accounts for 10 to 17% of the ergolines in the Argyreia products analysed. A uterine stimulant principle had been described in ololiuqui extracts as early as 1965. Nobody has ever quantified the abortifacient risk at recreational doses, nor calculated the dose of ergometrine actually ingested in a typical intake in order to compare it with obstetric doses. The absence of quantification does not lessen the contraindication. Breastfeeding follows the same rule.

3. Vascular risk, established as pharmacology and not as a case series. What holds: ergine is vasoconstrictive in vitro on bovine vessels, where it behaves as a partial adrenergic and 5-HT2 agonist; ergonovine is used in human clinical practice as a provocation agent for coronary spasm; raised blood pressure is among the autonomic effects of the halted human study; a case of PRES with hypertensive encephalopathy and status epilepticus is published. What does not hold: no human case of ergotism, of limb ischaemia or of gangrene after ingestion of morning glory seeds has been found in the indexed literature, and the PRES case is a single case in which LSA was not assayed, in which the species is not named, and in which the authors write only that the picture strongly suggests the attribution. PRES is a syndrome of hypertension and endothelial dysfunction, it is not ergotism. Populations to exclude, by extrapolation from the pharmacology of ergonovine and not from published cases: history of coronary spasm or of vasospastic angina, ischaemic heart disease, uncontrolled hypertension, peripheral arterial disease, Raynaud's phenomenon, history of stroke.

4. Vomiting while sedated is the most commonplace and the most serious chain. Gastrointestinal disturbance is among the most consistently reported effects, and no source read identifies the molecule responsible: neither the ergolines, nor the purgative resin glycosides of the Convolvulaceae, nor the seed coat material has been implicated by a measurement. Chanoclavine, the only ergoline of the seed documented at an emetic receptor, is a 5-HT3A antagonist that its authors propose as a potential antiemetic: it would go the other way. The common act of scraping off the whitish layer of the seed coat is recommended by at least one harm reduction organisation with no source at all. The actionable point is elsewhere: vomiting is the rule here, and anything that sedates turns a vomit into an inhalation. In this wiki's batch of substances, the sedatives are dextromethorphan, diphenhydramine and nitrous oxide; the other reliable emetics are Hawaiian baby woodrose, nutmeg and Syrian rue. Alcohol belongs to the same list, and it is present in the only death associated with LSA. With these seeds, where vomiting is not the accident but the rule, there is nothing to administer: what protects is somebody who has taken nothing, and a body lying on its side.

5. The coatings: what is true, and which way the rule runs. Fungicidal seed treatment is a real and regulated practice: an ornamental seed may carry a product that is not to be eaten. In the United States, the EPA regulation under FIFRA, 40 CFR 153.155, requires a seed treatment product to contain an approved dye giving the seed an unnatural colour, except where food tolerances exist, and with three exemptions. The Federal Seed Act, 7 CFR 201.31a, for its part requires only labelling. The implication runs one way only: a seed of unnatural colour is a treated seed, but an uncoloured seed is not thereby untreated, because of those exemptions, and above all because this is an American rule, with no bearing on seeds bought in France or in the European Union. As for the story that circulates everywhere, including in a toxicology column intended for emergency physicians, according to which commercial seeds are commonly coated with methylmercury to deter abuse: it has no identifiable primary source, three mutually incompatible versions of the same account circulate, and alkyl mercurials in seed treatment have been banned in the United States since 1970, methylmercury being one of them. Above all, nausea and vomiting are expected even with untreated seeds, which makes the coating useless as an explanation. Which also counts against extraction: it concentrates all the ergolines, hence ergometrine and LAH along with ergine, and it removes the natural ceiling imposed by vomiting, in practice the main factor limiting the absorbed dose. The systematic review notes that forums give misleading information on extraction and ingestion methods.

6. The name on the packet is worth nothing, in both directions. The psychoactive material is I. tricolor, in particular the three cultivars named above. But in the only data set in which commercial lots were analysed, what predicted the presence of ergolines was the cultivar name and not the printed species name: a lot sold as I. purpurea Heavenly Blue was positive, two other I. purpurea lots negative. The nomenclature itself has been muddying the trail for a century: the blue-flowered horticultural plant was long called Ipomoea violacea, including by Schultes, whereas it is I. tricolor, so that an older source writing I. violacea is almost always speaking of the blue-flowered plant. The idea that I. violacea in the strict botanical sense is devoid of ergolines rests on a single 2025 article read in abstract only, whose sample size, concentrations and detection limits are unknown, and which contradicts a century of literature without replication: plausible, not established. What holds without it is that legislative texts frequently target the inert species while omitting the psychoactive one. The genus Ipomoea moreover includes hundreds of never-tested species and contains plants that are toxic by mechanisms foreign to the ergolines: I. carnea and its swainsonine, which produces a lysosomal storage disorder, I. asarifolia and I. muelleri, tremorgenic for livestock through indole-diterpenes, mouldy sweet potato and its pulmonary oedema factor. A clarification that must be stated in both directions: no indole-diterpene has ever been measured in I. tricolor, only a gene cluster exists in its symbiont. Horticultural morning glory seeds are not known to contain tremorgens.

7. Combinations, in order of solidity. Documented as real co-use, with no clinical outcome attached: cannabis, very frequent, and dextromethorphan. The only death associated with LSA involves Argyreia seeds, a fall, and the subject carried THC and ethanol: a confounded picture, not to be attributed to LSA alone. On Syrian rue and the MAOIs, one must be exact in both directions. No published case of an interaction between LSA and an MAOI exists, and LSA is an ergoline, not a simple tryptamine: nobody has ever established that it is a substrate of monoamine oxidase, so that the usual caution, which comes from the tryptamine analogy, does not have the foundation attributed to it. That is not a reason to reassure: the class evidence goes the other way and it is published, the ergot alkaloids are serotonergic agonists and dihydroergotamine can produce a serotonin syndrome in humans, and the seed contains, as a matter of measurement, at least in one Argyreia lot, methysergide and methylergometrine. The foundation is mechanistic, not empirical. For triptans and the ergot medicines, the rule of non-association within a 24 hour interval is established, and applying it to the seeds is a reasonable extrapolation that must be presented as such; it is complicated by the fact that some users take these very seeds against cluster headache, an indication for the triptans. No case of serotonin syndrome attributed to LSA exists, and two independent searches confirm it: the absence is real, it is not a query defect. The warning about lithium that circulates on general-public sites has no primary source. There remains the seizure, and for this plant it is stated in the negative. The only intensive care picture published in this corpus, the Legriel 2008 PRES, includes status epilepticus, but the article is indexed under the molecule, LSA was not assayed in it and the species is not named in it: the morning glory therefore carries no convulsive signal of its own, and it is among the four substances of the corpus that carry none. What concerns it all the same is borrowed, and it is the only thing to retain here: ten of the fourteen substances of this corpus carry one, and the seed is not among them, but it is rarely taken alone, and what is added to it very often comes from the ten.

8. Signs requiring medical help, and what to say once there. Chest pain. Severe, unusual and persistent headache. Seizure. Confusion associated with raised blood pressure, a picture suggestive of PRES. Cold, pale or painful extremities. Intractable vomiting with inability to take fluids. Marked tremor and incoordination. Suicidal ideation, even fleeting, in the person or in what they say. Ergine appears in no routine assay and a negative screen excludes nothing: saying what was taken and keeping the packet with its cultivar label materially changes the management, all the more so as the printed species name may be wrong.

Sources

• A. Paulke et al. (2013), receptor profiling of LSA, J Ethnopharmacol (all the pKi values; conclusion that LSA is not an LSD-type psychedelic).
• C.M. Hazel and D.G. Panaccione (2025), Periglandula clandestina, symbiont of Ipomoea tricolor, Mycologia (fungal origin, 13 seed coats out of 37, tebuconazole, three ergolines, identical symbiont in the three cultivars).
• J. Nowak et al. (2016), ergolines of morning glory cultivars, Anal Bioanal Chem (seed weight, 3 positive lots out of 5, variability between single seeds, rearrangement of LAH, factor of 12 between plant and seed).
• W.T. Beaulieu et al. (2021), ergoline diversification and heritable symbionts, Commun Biol (67% of the herbarium samples; ergine as a hydrolysis product of LAH) and (2015), Periglandula in eight hosts, Mycologia (15.8 to 3223.0 µg/g between species).
• U. Steiner et al. (2011), erection of the genus Periglandula, Mycologia.
• P.S.C.C. Castro et al. (2025), systematic review of LSA, Pharmacy (Basel) (17 studies, 219 patients, 12.3% receiving medical care, no overdose recorded).
• A. Paulke et al. (2012), LSA in serum and urine after ingestion of Argyreia seeds, Anal Bioanal Chem (study halted for toxicity); (2014), ergoline profiles of two products, Forensic Sci Int and (2015), alkaloid composition, Forensic Sci Int (10 to 17% ergometrine; methysergide and methylergometrine in one lot).
• H.B. Klinke et al. (2010), two poisonings by Hawaiian baby woodrose seeds, Forensic Sci Int (death by fall, THC and ethanol present).
• S. Legriel et al. (2008), PRES with status epilepticus attributed to LSA, Neurocrit Care (LSA not assayed, species not named).
• J.W. Oliver et al. (1993), lysergamide-induced vasoconstriction, J Anim Sci and W. Auch-Schwelk et al. (2000), ergonovine contractions in human coronary artery, J Cardiovasc Pharmacol.
• S. Sueda et al. (2017), pharmacological spasm provocation testing, J Cardiol (clinical use of ergonovine; the class I there is a recommendation by the authors, not the state of the guidelines).
• S. Eom et al. (2021), regulation of 5-HT3A by chanoclavine and ergonovine, Molecules (IC50 107.2 µM; potential antiemetic).
• G.R. Juszczak and A.H. Swiergiel (2013), recreational use of D-lysergamide in Poland, J Psychoactive Drugs (suicidal ideation in two subjects; co-use of cannabis and dextromethorphan). Abstract only read, the tables of doses by species could not be reached.
• E. Sobanski et al. (2021), induced psychosis, Z Kinder Jugendpsychiatr Psychother and C. Ponte and M. Lapeyre-Mestre (2017), psychoactive effects of a legal high, Therapie (4 addictovigilance cases, 2 hospitalisations).
• S. Cohen (1964), suicide after ingestion of morning glory seeds, Am J Psychiatry. Record only, not read.
• S.T. Lee et al. (2017), indole-diterpenes of I. asarifolia and I. muelleri, J Agric Food Chem and K. Ikeda et al. (2003), alkaloids of I. carnea, J Agric Food Chem (toxicities of the genus foreign to the ergolines).
• US EPA, 40 CFR 153.155, dye requirement for seed treatment products, with three exemptions; USDA, 7 CFR 201.31a, labelling requirements for treated seed. Texts read. American rules, with no bearing in France or in the European Union.
• ICEERS, Ipomoea violacea: Basic Info (harm reduction organisation; ranges with no primary citation, seed count arithmetically wrong, active dose of 1 to 2 mg of ergine contradicted by Isbell).
• Erowid, Morning Glory Vault (corpus of user reports, to be read on site; its licence forbids reproduction). Not consulted, the server returned 403.
• H. Isbell and C.W. Gorodetzky (1966), effect of ololiuqui alkaloids in man, Psychopharmacologia 8(5):331-9, DOI. Text not read: no abstract in PubMed, second-hand abstract obtained through a search engine, corroborated only by the MeSH indexing of the record. Most important human source on this page.
• I. Mercurio et al. (2017), GC/MS analysis of commercial morning glory seeds, Egyptian Journal of Forensic Sciences 7:29, DOI. Full abstract read through a third-party reproduction, full text not read, number of samples unknown. The two content figures are carried over here; the authors' conclusion about doses capable of producing hallucinogenic effects is not, it is a claim without human data.
• E. Greco et al. (2025), analytical identification of the ergolines in Ipomoea, Forensic Chemistry 45:100686, DOI. Abstract only read, sample size, concentrations and detection limits unknown.
• A.L. Ingram (1964), morning glory seed reaction, JAMA 190:1133-4, DOI. Not read, available as PDF only. Probable original source of the medical statement about fungicidal coatings.
• W.B. Rice and K. Genest (1965), acute toxicity of morning glory seed extracts in mice, Nature 207:302-3, DOI. Not read, no abstract in the database. The only existing acute toxicity study on this material, and it is missing from any claim of non-lethality.
• M. Borsutzky et al. (2002), psychopharmacological effects of Argyreia seeds, Nervenarzt 73(9):892-6, DOI. English abstract only read, German-language article not read. A reasoned position, to be given as such.