Every season somebody in this forum asks the same question in a slightly different way. Are autos good enough yet? The honest answer is that “good enough” depends on what you are optimising. The interesting part of the answer changed in the last few years. Autoflowers did not suddenly get better. Researchers started mapping what the autoflower trait actually is.
This post picks up where the earlier post on regular, feminized and autoflowering seeds left off. That one covered what each seed type is. This one is about whether the gap has closed.
What was actually mapped
For about two decades “autoflowering” was a marketing word backed by breeder anecdote. Joint Doctor’s LowRyder work in the early 2000s crossed Cannabis ruderalis material into Indica and Sativa lines. It produced plants that flowered on age rather than on daylength. Nobody knew which gene was responsible.
A 2022 mapping study in hemp material (Toth et al., Frontiers in Plant Science) identified two major-effect flowering-time loci. Autoflower1 behaves as a simple recessive trait. In F2 populations the phenotypes split about 3:1, so roughly 25% of plants flowered under non-inductive long days (16 hours light / 8 hours dark). The genotypes underneath segregate 1:2:1, but heterozygous plants still look photoperiod-sensitive under long days. Candidate genes at the locus include a nuclear transcription factor Y subunit (NFYB1), an APETALA 2 floral homeotic protein and the pseudo-response regulator PRR37. A second locus, Early1, mapped to a casein kinase 1-like protein homologous to the rice gene Heading date 16. In field trials across three seasons it made a photoperiod-sensitive cultivar flower 2–4 weeks earlier.
Later genomic work on landrace collections has proposed more loci for photoperiod insensitivity. We could not verify the details well enough to name them here.
The finding that matters when you buy seeds
In their discussion, the 2022 authors report that “several commercially marketed cultivar populations from multiple sources were segregating for Autoflower1”, “which may have resulted in poor cultivar performance for growers due to variation in photoperiod sensitivity.”
That offers one plausible explanation for a complaint this forum sees constantly: half my pack finished in nine weeks and half ran thirteen, same tent, same feed. One likely cause is a seed line that is still segregating, whether at Autoflower1, at Early1 or at other flowering-time loci. It is not the only possible cause. Stress, light schedule and other genes can also stretch or split finishing times.
So the question to put to a seed vendor is not just “is this an auto?” It is: “Does every plant in this line autoflower (is it true-breeding), and how uniform is the finishing time?” Any plant that truly autoflowers is already homozygous recessive at Autoflower1. What you need to know is whether the whole line breeds true and finishes evenly. A vendor who cannot answer is telling you something.
The ceiling that has not moved
Genetics improved. Physics did not.
One controlled indoor trial (Rodriguez-Morrison, Llewellyn & Zheng, 2021, Frontiers in Plant Science) grew a photoperiod cultivar at canopy light levels from about 120 to 1,800 µmol·m⁻²·s⁻¹ PPFD. Flowering ran for 81 days at 12/12. Dry inflorescence yield rose linearly across the whole range, from about 116 to 519 g/m² (roughly 4.5-fold), with no sign of saturation. Cannabinoid concentration did not change with light level. Total cannabinoid yield rose only because biomass rose.
That is one trial on one cultivar, but it points in a clear direction. Concentration is largely set by genetics. Yield is largely set by how much light the canopy captures. They are different levers.
Now apply that to an autoflower. Its life cycle is typically about 8–14 weeks from seed. You cannot extend the vegetative phase to fill a canopy that is still half empty when flowering starts. That caps the total light each plant intercepts per cycle, and so caps grams per square metre per cycle. Breeding can shift the clock, but it cannot remove the cap. The limit comes from arithmetic, not only from genetics.
Per square metre per cycle, a well-run photoperiod crop generally wins. Per square metre per year, an autoflower schedule can win. Shorter cycles allow more cycles per year, and auto cycles do not need a separate vegetative room.
What about potency?
We could not find a peer-reviewed head-to-head comparison of cannabinoid content in modern autoflower and photoperiod cultivars grown under matched conditions. Anyone quoting you a percentage gap is quoting a seed catalogue or a marketing page, not a trial. We are not going to print a number we cannot source.
Here is what can be said about the mechanism. The historical gap came from ruderalis introgression, and ruderalis has very low cannabinoid content. Repeated backcrossing to high-cannabinoid parents while selecting for the recessive autoflower locus removes that background. There is no physical reason a fixed autoflower line cannot match its photoperiod parent in concentration. Whether a particular commercial line has done so is a question for a lab report on that line, not for a general claim.
Autoflowers in Thailand
Legal note. In Thailand, growing cannabis requires the applicable permit or notification. Cannabis flower is a controlled herb under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis) B.E. 2568. In Vietnam, Malaysia and other Southeast Asian countries, cannabis cultivation is illegal, often with severe penalties, and this article does not encourage it.
Daylength. Most grow guides are written for 45–55°N. Bangkok sits at about 13.7°N and Chiang Mai at about 18.8°N, where day length changes only modestly through the year. The posts on photoperiod plants in the tropics and photoperiod genetics for our latitudes explain why that complicates outdoor photoperiod growing. The short version for autos is that they ignore day length.
Rainy-season timing. A line with a known cycle length gives you a schedule you can plan against the monsoon calendar. You can count back from the week the humidity becomes unmanageable and know whether you will finish.
Light-deprivation setups. In a permitted light-dep greenhouse, autos are much less sensitive to a tarp failure. A light leak during the dark period can stress or delay a photoperiod crop, while an auto largely ignores it.
Beginners. Light leaks are a common indoor problem for new growers, and autoflowers largely remove that risk. They bring other risks, though, covered in the next section.
Where they are the wrong choice
Topic 1326 covers this in detail. In short, autoflowers generally do not make useful mothers or clones. They have little time to recover from early stress, and they tolerate only light, early training. Plant into the final container, feed lightly at the start and treat 20/4 as a convention, not a proven optimum. Above all, buy a line that breeds true.
The short answer
For potency, the gap is unmeasured. There is no physical reason for one, but there is also no published head-to-head comparison, so buy on lab reports rather than claims. For yield per cycle, a gap remains, and it comes mostly from arithmetic rather than genetics. For yield per year, per unit of grower attention and per unit of risk in a tropical wet season, autoflowers are frequently the better instrument.
What has genuinely improved since the LowRyder era is not the ceiling. We can now name the locus. That means a breeder can test for it and fix it deliberately instead of stumbling into it. It also means you have a specific, answerable question to ask before you buy.
Sources: Toth JA, Stack GM, Carlson CH, Smart LB, “Identification and mapping of major-effect flowering time loci Autoflower1 and Early1 in Cannabis sativa L.”, Frontiers in Plant Science 13:991680 (2022), doi:10.3389/fpls.2022.991680; Rodriguez-Morrison V, Llewellyn D, Zheng Y, “Cannabis yield, potency, and leaf photosynthesis respond differently to increasing light levels in an indoor environment”, Frontiers in Plant Science 12:646020 (2021), doi:10.3389/fpls.2021.646020.