Short answer. Read a seed listing for what you can check: the seed type (feminised, regular or autoflower), the parents and generation, the flowering time and how it's counted, and evidence from real grow journals, herm reports and documented lineage. Distrust flower potency above about 35 % total cannabinoids, effect promises built on "indica" and "sativa", and famous names with no records. Buy a small pack from a seller with documented lines, and check that buying and importing seed is legal where you live.
Buyer beware. Seeds don't come with a THC number; plants grown from them do, and only after a grow and a lab test. "Our seeds hit 42 %" is a promise the plant can't keep. See buyer beware: potency labels and other claims.
What do you need before you shop?
You need a few decisions made and a way to check claims.
- [ ] Your legal check done: home growing, plant limit, and seed purchase and import rules
- [ ] Your space and timeline: photoperiod (more control) or autoflower (faster) (autoflower first grow, photoperiod calendar)
- [ ] A plant count within your legal limit; for a first run, one to four
- [ ] A short list of 2–3 candidate lines, each with independent journals you have read
- [ ] A place to store seeds cool, dark and dry: an airtight container in the fridge at about 5 °C (41 °F) (seed viability and storage)
- [ ] A grow journal ready to record the pack, batch and GB-ID
What does each part of a seed listing tell you?
Each part of a listing is either checkable or marketing. Weight what you can check.
| Listing field | What it really tells you | What to look for |
|---|---|---|
| Seed type | Feminised: nearly all female. Regular: about half males. Autoflower: flowers by age | Matches your plan (seed types) |
| Parents | The cross, if honest | Both parents named; selection or cut named; not just a famous name |
| Generation | F1, F2, IBL, S1, backcross | Stated plainly; S1 on S1 is a warning sign for herm risk |
| Flowering time | Usually counted from first pistils, not the flip | A range (e.g. 8–10 weeks), and add 10–14 days from the flip |
| Height, yield | Grown in the breeder's room, not yours | Ranges, with conditions; treat as rough |
| THC / "potency" | A test of one sample, if any | A range from more than one lab; nothing above ~30 % THC or ~35 % total cannabinoids |
| Terpene or flavour notes | Descriptions | Pleasant to read; not checkable until you grow it |
| "Indica / sativa", effect claims | Almost nothing | Treat as plant shape at most |
| Journals and reports | How it really grew for others | Several independent grows, including problems and herm reports |
| Lineage evidence | How well the parentage is documented | GenetixBay L0–L4 and a GB-ID |
Why are potency claims above 35 % not credible?
Flower labelled above about 35 % total cannabinoids is not credible, and that is a GenetixBay house position. The founder credits Dr. Dedi Meiri with stating this limit at a seminar, with publication pending [3]. It holds from first principles: cannabinoids are made only in trichome heads [4], and most of a bud's dry mass is bract, stem and leaf. A generous estimate gives roughly 30–32 % total THC at most. A 42 % label would need about 48 % THCA in dry bud (42 / 0.877), a "bud" that is mostly loose resin heads.
The data agree. Across 215,285 Washington tests, the 99th percentile was 27.0–31.8 % depending on the lab, and results varied systematically between labs [1]. Retail flower labels often overstate THC when retested [2]. Most flower people believe is mid-20s to 30s really tests mid-teens to low 20s. Growing even 30 % is a real feat. See label inflation and lab shopping.
Do "indica" and "sativa" mean anything on a listing?
As a guide to effect or species, no. That is a house position backed by evidence: nearly everything sold is a hybrid within one species, Cannabis sativa L. Genetic analysis found the labels don't track genetic structure [5], and commercial chemistry clusters into a few groups that the labels don't predict [6]. At most, "indica" and "sativa" describe plant shape: broad or narrow leaflets, structure and flowering time. See indica, sativa and ruderalis.
Why does the same name not mean the same plant?
The same name often doesn't mean the same genetics, because names are copied, reused and attached to different cuts. Genetic testing found samples sold under one strain name that were not closely related [7]. And many "new" names are the same cut renamed, which narrows what's actually on offer (same cut, a thousand names). Buy the documented line, not the name.
How do you choose and buy seeds, step by step?
Choose and buy in this order.
- Do the legal check for growing, plant numbers, buying and importing seed.
- Decide photoperiod or autoflower from your space and time, and feminised or regular from your goal (flower or breeding).
- Shortlist 2–3 lines with stated parents and generation.
- Read at least three independent journals for each. Note real flowering time from the flip, height, problems and any intersex flowers.
- Check for herm reports. An occasional outlier happens; a pattern is poor breeding. Avoid lines with repeated reports (telling sex and spotting herms).
- Check the potency claim. Discard anything promising more than ~30 % THC or ~35 % total cannabinoids, or "guaranteed" effects.
- Check how flowering time is counted. Add 10–14 days from the flip if it counts from pistils [8].
- Prefer documented lineage: a GB-ID and an evidence level of L2 or higher, published ranges and lab names (documenting lineage evidence).
- Buy a small pack first, for example 3–5 seeds, and grow it before buying more.
- Store unused seed airtight, dark and cold, and log the pack, batch and purchase date.
What separates a breeder from a pollen chucker?
A breeder selects, tests and documents; a pollen chucker crosses two popular plants and sells the seed untested. The difference shows on a listing: parents, generation, selection work, tester grows and published herm results versus a hype name and a headline number. Hermaphroditism has real consequences for seed and progeny [9], and the GenetixBay position is that herm-prone lines are mainly the product of bad breeding. Late-flower "bananas" are not normal: they are herms. See breeders vs pollen chuckers.
How will you know you bought well?
You'll know you bought well when:
- Germination is high, 80–100 % for fresh seed (germinating seeds).
- Plants from one pack look broadly similar in structure and timing, as the listing's generation suggests.
- Flowering lands within the stated range plus the 10–14 day transition.
- No intersex flowers appear under a clean, leak-free environment.
- Your results match the independent journals you read before buying.
What mistakes do beginners make buying seeds?
Beginners mostly buy the headline instead of the evidence.
- Buying the highest THC number on the page.
- Choosing by "indica for sleep, sativa for energy".
- Buying a famous name with no parents, generation or journals.
- Expecting harvest at the stated weeks from the flip.
- Buying regular seed by accident for a flower-only grow, or feminised seed when you want to breed.
- Buying large packs of an untested line.
- Ignoring seed import rules.
- Storing seed warm in a drawer for years.
Evidence
- Lab-to-lab variation and the high end of real flower data: Established [1].
- Retail labels overstating THC: Supported [2].
- The ~35 % ceiling: house position (Meiri, seminar remarks; publication pending) [3], consistent with [1][2][4].
- Indica/sativa labels don't track genetics or chemistry: Supported [5][6].
- Strain names unreliable: Supported [7].
- Flowering time counted from pistils; 10–14 day transition: Supported [8]; house position.
- Herms mainly from bad breeding: house position, interpreting [9].
