| Molecule | CAS | Levels | R/G/N/A | Evidence summary | Sources |
|---|---|---|---|---|---|
| (R)-(−)-carvone monoterpene ketone | 6485-40-1 ✓ | RGNA | 3/1/3/2 | Certain: OR1A1 responds to (R)-carvone, and the structural basis for enantiomer and ligand discrimination within one receptor is now resolved (distinct subpockets). CAUTION: the two EC50 values reported for OR1A1 x (R)-carvone differ by roughly three orders of magnitude between labs — a concrete… | 10.1038/s41586-024-08126-0 · 10.1073/pnas.1804106115 · Howard 2009 |
| L-menthol monoterpene alcohol — trigeminal agent (TRPM8), dual olfactory+trigeminal | 2216-51-5 ✓ | RGNA | 4/1/2/1 | Certain: OR1A1 responds to L-menthol at ~16 uM and the structural basis is resolved (N109 3x37 H-bond, TM5 subpocket). Notable: OR8K3 is a SEGREGATING PSEUDOGENE with (+)-menthol as agonist — some humans simply lack this receptor — but we found no measurement of what that does to menthol perception.… | 10.1038/s41586-024-08126-0 · 10.1038/nn.3598 · 10.1371/journal.pone.0121091 |
| linalool terpene alcohol (monoterpenoid) | 78-70-6 ✓ | RGNA | 3/2/2/4 | Two human receptors (OR1C1, OR2A25) validated for linalool with concentration-specific genotype effects. No EC50, no enantiomer specification, no allele frequencies retrieved. Number of human neural studies that DO use linalool as a single substance: 1. Number of studies reporting the effect of linalool ON ITS OWN: 0. No… | 10.1073/pnas.1804106115 · Iravani 2021 · Grabenhorst 2007 |
| trimethylamine tertiary aliphatic amine (trace amine) — TAAR5 ligand (not via classical ORs) | 75-50-3 ✓ | RGNA | 4/2/2/3 | Certain: hTAAR5 is activated by TMA (EC50 116 uM) and DMEA (169 uM) and by nothing else among 42 amines. Certain: TAAR5 p.Ser95Pro reduces fish-odour intensity and unpleasantness in 9,122 Icelanders, replicated. IMPORTANT: Wallrabenstein 2013 is a NEGATIVE study for the genetics — it found NO… | 10.1371/journal.pone.0054950 · hal-04095283 (HAL) / ECRO 2022 poster Fri-P2-104 · 10.1016/j.cub.2020.09.012 |
| vanillin phenolic aldehyde (methoxyphenol) | 121-33-5 ✓ | RGNA | 3/2/3/3 | Certain: vanillin activates OR10G4 in vitro but OR10G4 genotype does NOT affect vanillin perception; vanillin activates many ORs. This is the cleanest published demonstration that in vitro deorphanization does not imply perceptual relevance. Also: vanillin pleasantness is the phenotype most… | 10.1038/nn.3598 · 10.3389/fphar.2015.00035 · 10.1073/pnas.1804106115 |
| 2-ethylfenchol bicyclic terpene alcohol | NR | RG-A | 1/1/0/1 | The best-characterised NEW pair from the repertoire-wide survey: OR11A1 haplotype frequencies are known, the in vitro effect matches the perceptual effect, and the neighbouring ORs were excluded. But the EC50 is ~10 mM, i.e. extremely weak — a caution that in vitro potency and perceptual relevance… | 10.1073/pnas.1804106115 |
| androstadienone 16-androstene steroid — C19 steroid (androstene), non-hormonal derivative of testosterone | 4075-07-4 ✓ | RGN- | 2/3/5/0 | Androstadienone is the clearer OR7D4 case for VALENCE (not only intensity). No EC50 retrieved. The molecule with the RICHEST human neural profile, and the ONLY molecule with a dose–response curve in human olfactory fMRI (Burke 2012: 3 molar concentrations). Hypothalamic activation replicated across 2 independent research groups… | 10.1038/nature06162 · 10.1073/pnas.1804106115 · 10.1073/pnas.0407998102 |
| androstenone 16-androstene steroid (pheromone-like semen/urine volatile; boar taint component) | 18339-16-7 ✓ | RG-A | 2/2/0/1 | Certain: OR7D4 RT/WM alters androstenone intensity perception and this transfers to real food. Uncertain: population allele frequencies, and whether the valence shift is a direct receptor effect or a consequence of lower intensity. | 10.1038/nature06162 · 10.1073/pnas.1804106115 · 10.1371/journal.pone.0035259 |
| cis-3-hexen-1-ol C6 green-leaf volatile (unsaturated primary alcohol) | 928-96-1 ✓ | RG-A | 3/1/0/1 | Certain: OR2J3 carries two variants that together abolish in vitro response and the haplotype explains 26.4% of threshold variance in 52 people. IMPORTANT CAVEAT: a second independent lab could not reproduce the OR2J3-leaf-alcohol response (10.1523/JNEUROSCI.3259-15.2016), so the receptor… | 10.1093/chemse/bjs049 · 10.1523/JNEUROSCI.3259-15.2016 |
| citral acyclic monoterpene aldehyde (in fact a geranial/neral mixture) | 5392-40-5 | R-NA | 1/0/3/1 | The standard ‘pleasant’ control molecule in human valence imaging, with 2 attributable studies. But technically speaking, citral is a geranial/neral mixture, not a pure substance. Only one human study (Pei 2024). Serious problem: citral was the arm with the highest ODOUR INTENSITY and the highest FAMILIARITY… | 10.1038/nn1001 · Winston 2005 · Howard 2009 |
| guaiacol methoxyphenol (lignin pyrolysis product; smoke/whisky/coffee volatile) | 90-05-1 ✓ | RG-A | 4/1/0/1 | Certain: OR10G4 haplotypes change guaiacol potency ~100-fold in vitro and explain 15.4% of perceived-intensity variance and 10.3% of valence variance in 308 people, with loss of function making guaiacol MORE pleasant. Uncertain: which residue combination causes the loss; population-level haplotype… | 10.1038/nn.3598 · 10.1073/pnas.1804106115 |
| sulfur compound in urine after eating asparagus volatile organosulfur mixture | NR | RG-A | 3/1/0/1 | Certain: a single OR2 cluster at 1q44 is robustly associated (p down to 4e-44) with self-reported asparagus-urine anosmia in 6,909 European-Americans, and 'anosmia' is the MAJORITY phenotype (58-62%). NOT established: which receptor, which ligand, or any EC50 — neither OR2M7 nor OR2L3 has ever been… | 10.1136/bmj.i6071 · 10.1007/s00018-019-03279-y |
| indole nitrogen-containing heterocycle (benzopyrrole) | 120-72-9 | R-NA | 3/0/3/1 | One of the few molecules that IS presented on its own with an explicit concentration in human olfactory fMRI (Grabenhorst 2007, 6%). Important because it is the textbook example of concentration-dependent valence reversal. CERTAIN: detection threshold 0.30 ppb (one methodologically consistent source). CERTAIN: a 4-fold increase in concentration reverses the valence… | 10.1016/j.cub.2020.04.086 (Pfister et al. 2020 Curr Biol 30(13):2574-2587.e6; read via the bioRxiv preprint 10.1101/760033v2) · as above · Grabenhorst 2007 |
| limonene monoterpene hydrocarbon; also a TRP ligand → has a trigeminal component at high concentration | 138-86-3 (racemic); 5989-27-5 (R)-(+)/D; 5989-54-8 (S)-(-)/L | R-NA | 1/0/3/2 | Number of human fMRI/PET studies reporting the effect of limonene ON ITS OWN: 0. Number of EEG studies: 1, metadata only, no n, no concentration, no statistics, and a conclusion based on reverse inference. NOT CERTAIN AT THE MOST BASIC LEVEL. Two studies using the SAME isomer ((R)-(+)/D) reached OPPOSITE conclusions: Heuberger… | Howard 2009 · Grabenhorst 2007 · 10.1002/ffj.70133 |
| octanal medium-chain aliphatic aldehyde | 124-13-0 ✓ | RG-A | 4/1/0/1 | Certain (as of 2026): OR6A2 is now structurally deorphanized with octanal bound, via a previously unknown reversible covalent Schiff-base mechanism at K4.60. Certain: rs72921001 in the 11p15 OR cluster associates with self-reported cilantro soapiness at OR=0.81. NOT established: that rs72921001… | 10.1016/j.cell.2025.12.017 · 10.1186/2044-7248-1-22 |
| pentanoate short-chain fatty acid (straight-chain C5 carboxylic acid) | 109-52-4 ✓ | RGN- | 2/2/3/0 | Fills the OR51E1 gap left open by Billesbolle 2023, which inferred but did not measure OR51E1's ligand: OR51E1 prefers pentanoate (C5) at ~23 uM. No human genetic or perceptual data. The ‘unpleasant’ molecule most used in human olfactory imaging. Two studies allow separate attribution… | 10.1038/s41586-024-08126-0 · Menashe 2007 — see a1 · 10.1038/nn1001 |
| trans-anethole phenylpropanoid (methoxy-propenylbenzene) | 4180-23-8 ✓ | RG-A | 2/1/0/1 | Certain: a common OR6C70 missense variant (57% in East Asians) increases both intensity and pleasantness of licorice. Not established: OR6C70 has not been deorphanized in vitro with trans-anethole; chromosome 12 carries ~300 correlated OR variants, so the causal gene is not resolved. | 10.1016/j.cub.2020.09.012 |
| trans-cinnamaldehyde aromatic aldehyde (phenylpropanoid) — trigeminal agent (TRPA1) | 14371-10-9 ✓ | RGN- | 3/1/2/0 | The notable claim here is conceptual: this is the first large study to show that a common variant affects odour NAMING/identification, a cognitive-level phenotype, and not only peripheral sensitivity. The variant is intergenic, so the mechanism is unknown. OR2B11 is a separate, in-vitro-confirmed… | 10.1038/nn.3598 · 10.1016/j.cub.2020.09.012 · 10.1371/journal.pone.0121091 |
| β-ionone cyclic terpenoid ketone (carotenoid degradation product) | 14901-07-6 ✓ | RG-A | 1/2/0/1 | Certain: rs6591536/N183D in OR5A1 is the strongest near-Mendelian olfactory trait known in humans (96%+ classification accuracy) and the insensitive allele is most common in East Asians. Uncertain: no EC50 retrieved; the '95% of variance' figure is a loose claim with no variance statistic attached;… | 10.1016/j.cub.2013.07.030 · 10.3390/brainsci11111405 · 10.1016/j.foodres.2014.02.038 |
| (S)-(−)-citronellol and (S)-(−)-citronellal acyclic monoterpenoid alcohol / aldehyde | NR | RG-- | 2/1/0/0 | An early (2007) structure-function study of two human OR paralogs, notable for the historical baseline it records: at the time, only 6 of 380 human ORs had known ligands (~1.6%). By 2024 the figure was ~12% (10.1093/chemse/bjae037) — roughly 20 years for a 10-point gain. | 10.1016/j.jsb.2007.04.013 |
| 1-butanol straight-chain C4 alcohol — the standard substance for olfactory threshold testing | 71-36-3 | R-N- | 1/0/3/0 | An important control substance that receives little attention. Burke 2012 gives an explicit molar concentration (0.01 mM) and a clear result: butanol activated the whole olfactory network (amygdala/insula/OFC bilaterally) but did NOT activate the hypothalamus. | 10.1371/journal.pone.0040993 · 10.1073/pnas.0407998102 · 10.1016/j.neulet.2020.135166 |
| 2-phenylethanethiol aromatic thiol | 4410-99-5 ✓ | RG-- | 2/1/0/0 | Two broadly tuned human ORs (OR2W1, OR1A1) with measured EC50s in the tens-to-hundreds of micromolar range, and opposite copper sensitivities. No human perceptual or population-genetic data. | 10.1007/s00018-019-03279-y |
| 2-phenylethanol aromatic alcohol | 60-12-8 | R-N- | 1/0/6/0 | The PURELY OLFACTORY molecule with the best human neural profile: >=6 independent studies, 1 paper with a complete MNI table and an explicit concentration (Ekanayake, 1% in mineral oil). The ‘purely olfactory’ criterion is demonstrated behaviourally: the nostril cannot be localised. human_affect EMPTY. a3 searched with many queries and… | 10.1093/chemse/bjae038 · 10.1002/hbm.26772 · 10.1371/journal.pone.0038358 |
| 3-mercapto-2-methylpentan-1-ol polyfunctional thiol (mercapto-alcohol) | 227456-27-1 ✓ | RG-- | 1/1/0/0 | Certain: OR2M3 is the highest-affinity human OR-odorant pair in this dataset (EC50 0.29-0.45 uM) and its response depends on a conserved CC/CSSH copper-binding motif (Cys202/Cys203 and Cys241/Ser242/Ser243/His244), 100% conserved across 23 human family-2 ORs and 42 mammalian orthologs. Certain:… | 10.1007/s00018-019-03279-y |
| bourgeonal aromatic aldehyde | 18127-01-0 ✓ | RG-- | 1/1/0/0 | OR10J5 validated for bourgeonal in 332 people. Bourgeonal is widely cited in the (contested) human-sperm-chemotaxis literature; that claim is unrelated to this nasal perceptual finding and should not be merged with it. | 10.1073/pnas.1804106115 |
| caproic acid short/medium-chain saturated fatty acid | 142-62-1 ✓ | RG-- | 2/2/0/0 | The clearest receptor-level demonstration in the human literature that a single odorant is encoded by different receptors at different concentrations — relevant to a4's concentration-dependent valence reversal brief. | 10.1073/pnas.1804106115 |
| carbon dioxide inorganic gas — the gold-standard purely trigeminal control substance | 124-38-9 | R-N- | 1/0/4/0 | The best purely trigeminal control substance available, with consistent data across PET, fMRI and an ALE of 9 studies/~170 people. The most important result: odourless CO2 still activates the piriform cortex with every method. | 10.1002/hbm.20573 · 10.1016/j.brainresrev.2009.11.001 · 10.1371/journal.pone.0121091 |
| estratetraenol C18 estrogen-like steroid | 1670-49-1 | R-N- | 1/0/2/0 | Markedly weaker than AND: no known human receptor, no dose–response curve, and the key results include an uncorrected value (P=0.013) and a sub-threshold conjunction (P=0.06). There is no evidence that EST exists at a meaningful physiological concentration in human sweat. | 10.1073/pnas.0407998102 · 10.1073/pnas.0600331103 |
| ethyl vanillin phenolic aldehyde | 121-32-4 ✓ | RG-- | 1/1/0/0 | Same pattern as vanillin: in vitro agonist of OR10G4, no perceptual consequence of OR10G4 genotype. | 10.1038/nn.3598 |
| eugenol phenolic phenylpropanoid; a TRPV1/TRPA1 ligand → does have a trigeminal component; used as a topical dental anaesthetic | 97-53-0 | R-N- | 1/0/3/0 | Appears in >=4 human neural studies, BUT NO study reports the effect of eugenol on its own. It is always pooled (ordinary odors / neutral odors) or a component of a mixture. | 10.1073/pnas.0407998102 · 10.1523/JNEUROSCI.1835-13.2013 · Grabenhorst 2007 |
| isoeugenol phenylpropanoid (methoxyphenol) | 97-54-1 ✓ | RG-- | 2/1/0/0 | The clearest published case of an OR variant shifting hedonic valence WITHOUT shifting intensity, which argues that valence is not merely a read-out of perceived strength. | 10.1073/pnas.1804106115 |
| muscone macrocyclic musk ketone (MCM) | 541-91-3 ✓ | RG-- | 5/1/0/0 | Certain: OR5AN1 is the principal human muscone receptor (EC50 12.5 uM) and deleting the single orthologous receptor in mouse raises the behavioural muscone threshold by 2-3 orders of magnitude while leaving eugenol detection intact — mechanistic proof that one OR can carry one odour. Missing… | 10.1523/JNEUROSCI.3259-15.2016 |
| musk ketone nitro musk (NM) | 81-14-1 ✓ | RG-- | 3/1/0/0 | Highest-affinity human musk-receptor pair documented here. No human variant or perceptual data. | 10.1523/JNEUROSCI.3259-15.2016 |
| musk xylol nitro musk (NM) | 81-15-2 ✓ | RG-- | 4/1/0/0 | Three human ORs assigned. The OR2J3 result directly conflicts with OR2J3's assignment to cis-3-hexen-1-ol. | 10.1523/JNEUROSCI.3259-15.2016 |
| propionate short-chain fatty acid (SCFA; gut-microbial fermentation product) | 79-09-4 ✓ | RG-- | 2/1/0/0 | Certain: the first cryo-EM structure of a human odorant receptor with its odorant, giving the complete recognition chemistry for a class I OR (ionic anchor R6x59 + two H-bonds + a 31 A3 pocket). NOT available: EC50 numbers in the accessible text, any human variant, any human perceptual data. | 10.1038/s41586-023-05798-y · 10.1038/s41586-024-08126-0 |
| 1,8-cineole monoterpenoid cyclic ether | 470-82-6 | ---A | 0/0/0/1 | There are NO data on 1,8-cineole ADMINISTERED as a single molecule in humans. The only evidence is a correlation of r~0.43-0.50 between SERUM cineole and subtraction scores in 20 people with NO control group, with cineole completely confounded with exposure duration (r=0.727). | 10.1177/2045125312436573 |
| alpha-pinene bicyclic monoterpene (hydrocarbon) | 80-56-8 (racemic); 7785-70-8 (+); 7785-26-4 (-) | ---A | 0/0/0/1 | There IS exactly one human study with the pure molecule: HF-HRV rose 46.8% and HR fell 2.8% in n=13 women, 90 seconds, one-tailed test, no effect size. In that very paper, 'relaxed' and 'natural' did NOT reach significance - i.e. subjective relaxation did not change even though HF rose 47%. NO sleep outcome of any kind in… | 10.1007/s10086-016-1576-1 |
| alpha-santalol sesquiterpene alcohol | 115-71-9 | ---A | 0/0/0/1 | One of the very few 'oil vs main constituent' pairs measured in parallel in humans. But none of the quantitative figures are traceable (full text closed, site blocking bots). | 10.1055/s-2006-941544 |
| beta-caryophyllene bicyclic sesquiterpene (hydrocarbon); CB2 agonist | 87-44-5 | ---A | 0/0/0/1 | Of the 3 molecules in Pei 2024, beta-caryophyllene has the strongest WITHIN-GROUP profile (the only one to lower total STAI, S-AI and T-AI, and the only one to raise happiness). But the BETWEEN-GROUP comparison against the solvent was NEGATIVE, n=12/group, and there is no effect size at all. | 10.1016/j.heliyon.2024.e38941 |
| cedrol sesquiterpene alcohol | 77-53-2 | ---A | 0/0/0/2 | The MOST CERTAIN item in the whole a3 set in MECHANISTIC terms: cedrol lowers blood pressure and raises BRS in humans, and the effect PERSISTS when olfaction is completely eliminated (laryngectomised people, delivery into the lower airway) and in people who have lost their sense of smell. => this is PULMONARY/SYSTEMIC PHARMACOLOGY, not an olfactory effect. NOT CERTAIN: the magnitude… | 10.1016/j.autneu.2003.08.002 · 10.1111/j.1365-2125.2007.02992.x |
| Hedione synthetic cyclopentanone ester; jasmonate agonist | 24851-98-7 (methyl dihydrojasmonate; a3 could NOT verify this against the original paper) | ---A | 0/0/0/1 | The BEST-reported single-molecule human study (with eta2p and d, a power analysis, a standardised TSST stressor, n=56) - and the result is NEGATIVE. This is the most important datum for calibrating expectations: when a single odour molecule is tested with relatively good methods, the effect disappears (eta2p = 0.00… | 10.1080/10253890.2021.1961128 |
| isovaleric acid branched-chain C5 carboxylic acid | 503-74-2 | --N- | 0/0/2/0 | The human neural profile comes from ONE study, and there isovaleric acid was used IN A MIXTURE with cheddar flavour. The value of the paper is not about the molecule but about showing that semantic context determines the brain response. | de Araujo 2005 · Torske 2022 |
| linalyl acetate terpene ester (monoterpenoid acetate) | 115-95-7 | ---A | 0/0/0/1 | This is one of the very few molecules with a human study that CONCENTRATION-MATCHES it against the corresponding essential oil. The hypotensive effect of the pure molecule is the clearest one (DBP p<0.01) and is an effect that the oil containing it does NOT have. But there is no effect size, and the between-group comparison for anxiety was negative. | 10.7586/jkbns.2021.23.4.267 |
| undecanal aliphatic aldehyde (C11) | 112-44-7 | R--- | 1/0/0/0 | Important as one of the very few antagonist pairs with data IN HUMANS. The original Spehr 2003/2004 papers need to be read. | Spehr et al. 2003, 2004; Brodin et al. 2009 (qua review Thomas-Danguin 2014, DOI 10.3389/fpsyg.2014.00504) |
| 1-p-menthene-8-thiol monoterpenoid thiol | 71159-90-5 | ---- | 0/0/0/0 | No traceable figures. The number ‘10^-4 ppb’ circulates widely in the industry literature but could NOT be verified against an original source. | — |
| civetone macrocyclic ketone (musk) | 542-46-1 | ---- | 0/0/0/0 | The ONLY quantitative data (from an INDUSTRY technical article, Perfumer & Flavorist, NOT peer-reviewed): civetone makes up 2.5-3.5% of natural civet; civet absolute is used mainly as a 3% alcoholic solution; tinctures are usually 5% (sometimes 50%); the Pasteur Institute Addis Ababa classification: >60% absolute = adulterated, 50-60% =… | — |
| decanal aliphatic aldehyde (C10) | 112-31-2 | ---- | 0/0/0/0 | Threshold only. | — |
| methyl mercaptan thiol | 74-93-1 | ---- | 0/0/0/0 | Used as a thiol benchmark with a verifiable threshold, in place of grapefruit mercaptan (the original paper could not be accessed). | — |
| nonanal aliphatic aldehyde (C9) | 124-19-6 | ---- | 0/0/0/0 | Threshold only. | — |
| p-cresol alkylphenol | 106-44-5 | ---- | 0/0/0/0 | Threshold only. No concentration-dependent valence data. | — |
| skatole benzopyrrole (aromatic N-heterocycle) | 83-34-1 | ---- | 0/0/0/0 | CERTAIN: threshold 5.6 ppt — ~54 TIMES MORE SENSITIVE THAN INDOLE despite differing by only one methyl group. This is the cleanest structure-activity figure in this database. NOTHING YET on concentration-dependent valence. | — |
| TMT thiazoline — a component of fox faeces, used as an innate fear-inducing substance in rodents | 54717-17-8 | ---- | 0/0/0/0 | NO human imaging or neurophysiological study using TMT WAS FOUND. The entire TMT literature is in rodents. | — |
Levels: R = human receptor · G = genetic variant · N = human neural · A = human affect. The R/G/N/A column gives the number of records with a real source at each level. ✓ = CAS number cross-checked against a chemical database.
Scientific reference material, not medical advice and not promotional material. All assessments of evidence quality are the compilers’ own. Figures are redrawn from published data.