Watch the episode · 4:24Every strain you have ever grown began with two parents and a wish list. Making a truly new strain takes years, not one lucky cross. Here it is, step by step. And the twist: when scientists DNA-tested strains bought under the same name, most of them had at least one sample that didn't match.
The wish list ▶ 0:29

Step one is not a seed at all. A breeder starts with a goal: a bigger yield, a certain aroma, faster flowering, resistance to mould. Then they look for two parents that, between them, carry those traits.

The mother is easy to judge: you can see and smell her flowers. The father is harder. Males don't make buds, so breeders judge them on vigour and structure, and best of all by growing out their daughters.
The cross: F1 ▶ 0:59

Step two: pollen from the chosen male is brushed onto the female's flowers. The seeds she makes are the first generation, the F1.
In maize, F1 seeds come from two inbred parents, so every plant is almost identical. Cannabis is different. It naturally cross-pollinates, so most plants carry two different versions of many genes, and even F1 plants can vary.
Source · Garcia-de Heer et al., Hortic. Res. (2026)
The shuffle: F2 ▶ 1:25

Then breeders do something strange: they cross the children with each other. Here is a real example. Cross a pure THC plant with a pure CBD plant and every F1 makes a mix of both. Cross two of those and the F2 splits: about one in four THC, two in four mixed, one in four CBD. That is exactly what researchers found in 2003.
Now do that for every trait on the wish list. Want three traits, each set by one gene like this, all locked in? On average only one F2 plant in 64 gets the full set, and most traits depend on many genes, so the real odds are worse.

That is why breeders grow hundreds of F2 plants. It is the pheno hunt from episode 6: grow big, compare everything, and keep only the few that tick every box.
Locking it in ▶ 2:17

Finding the one is easy; making it breed true takes years. The breeder selects the best, breeds them together, grows out their seeds and checks every plant. Then repeats. Each round, more genes become matching pairs and the offspring look more and more alike.
- 01Select
keep the best few
- 02Inbreed
cross them together
- 03Grow out
check every offspring
- 04Repeat
generation after generation
After about five or six generations the line breeds true: its seeds grow into plants that look like their parents. That is the IBL, the inbred line, from episode 2 on seed pack codes. At one crop a year outdoors, that is years of work.

One more tool is the backcross: cross the line back to one of the original parents to win back a trait that slipped away. That is the BX on a seed pack.
Why names mislead ▶ 3:03

So why do two packs with the same name grow so differently? Because that slow work is expensive. Many strains are sold after just one or two generations, then re-bred and renamed by others. In a study of 30 strains from shops in three US states, 26 had at least one sample that didn't match the rest.
The future may look more like maize. In 2026 an Australian team inbred cannabis for up to six generations in just 18 months, then crossed those lines. Their F1 hybrids were more uniform than the originals, and more vigorous.
Source · Garcia-de Heer et al., Hortic. Res. (2026)
| Quick cross | Stable strain | |
|---|---|---|
| Generations | one or two | about six of selection |
| Its seeds | vary a lot | grow true to type |
| Time | months | years |
Set a goal and pick two parents. Cross them to make the F1, cross the F1s and the F2 shuffles the traits. Hunt through hundreds and keep a few, then inbreed for about six generations until the line breeds true.
Sources
- de Meijer, E.P.M. et al. (2003). "The inheritance of chemical phenotype in Cannabis sativa L." Genetics 163(1): 335–346.
- Schwabe, A.L. & McGlaughlin, M.E. (2019). "Genetic tools weed out misconceptions of strain reliability in Cannabis sativa: implications for a budding industry." Journal of Cannabis Research 1: 7. doi:10.1186/s42238-019-0001-1
- Garcia-de Heer, L., Mieog, J., Burn, A., Nolan, M., Liu, L., Manansala-Siazon, S. & Kretzschmar, T. (2026). "Weeding out variability: a proof-of-concept for producing uniform F1 hybrid Cannabis sativa L. using single-seed descent." Horticulture Research 13: uhag038. doi:10.1093/hr/uhag038
Every claim in this article comes from the episode script, which is checked against these papers. Prices on Seedscanner are dated observations; this page is not growing or legal advice.





