Short answer. Switching to 12/12 starts floral induction. It doesn't start flowering. The shoot tip spends the next week or two rebuilding itself, and the first real pistil clusters at the top of the plant usually show about 8–14 days after the flip in controlled studies [1][4], with slow cultivars taking three weeks or more [4]. Breeders' flowering times often count from that point, not from the flip. So a breeder's "8 weeks" usually means harvest about 9.5–10 weeks after the flip. Count both: days since flip and days of flower.
House position: the flip isn't day 1 of flower. The founder's ruling: after the switch to 12/12 there's a transition, generally about 10–14 days, before flowering really begins, and most growers don't account for it. That's why most runs go "longer" than the breeder's stated time by about the same amount: breeder times often count from first visible pistils or "flower set", not from the flip. Count from the flip and add the transition. Jardin and GenetixBay journals track both "days since flip" and "days of flower". The evidence for and against is below.
The founder, on "how will I know when she's actually flowering?": "Talk to her." No, seriously: she'll tell you more than you bargained for.
What's happening after the flip
Cannabis is a short-day plant: it measures the uninterrupted night (see photoperiod and light leaks). Once the night is long enough, the shoot apex changes what it builds.
- Under long days the main shoot keeps making long internodes, big fan leaves and side shoots. Each node carries a pair of single flowers once the plant is old enough [1][2].
- Under short days the apex branches intensively, the internodes shrink and bracts stack into a compact inflorescence: the cola [1][3].
In one Israeli study, the apex of a cultivar showed bract and flower primordia under the microscope by about day 7–8 of short days, but visible pistils at the top of the main shoot didn't appear until day 12; another cultivar showed them by day 8 [1]. Full flowering, with inflorescences on main, second- and third-order branches, took until about day 22 [1]. The plant is busy from day one. It just isn't visibly flowering yet.
The switch itself is fast and reversible. A follow-up study found inflorescence onset needed at least three consecutive short days, short days had to continue to keep the inflorescence compact, and returning plants to long days raised gibberellin and auxin at the apex and pulled the inflorescence apart [3]. That's the biology of re-veg, and why a leaky dark period in the first two weeks costs you time.
Preflowers are not the start of flower
The single flowers that appear at the nodes in veg ("preflowers") are not induced by the flip. Spitzer-Rimon and colleagues concluded that "induction of solitary flowers is age-dependent and controlled by internal signals, but not by photoperiod" [1]. In seedlings the juvenile-to-adult shift happened autonomously; in the cultivar studied, the first pair of single flowers formed at the 7th node [2]. Alter et al. confirm that solitary-flower initiation in seedlings is photoperiod-independent [3].
| Solitary flowers ("preflowers") | Induced inflorescence (real flowering) | |
|---|---|---|
| Trigger | Plant age, internal signals [1][2] | Long uninterrupted nights [1][3][4] |
| Appears under 18/6 or 24/0? | Yes | No (photoperiod types) |
| Where | One pair per node, in leaf axils | Clustered at the top of the main shoot and branch tips |
| What it tells you | Sex, and that the plant is mature | That flowering has started: day 1 of flower |
The research definition of flowering onset is the one to use: at least three pairs of pistils visible at the top of the main shoot [1][7]. Zhang et al. used the equivalent: forked stigmas protruding from bracts at the apical meristem [4].
How long the transition takes: the data
| Study | Plants and schedule | Time from short days to flowering onset |
|---|---|---|
| Spitzer-Rimon et al. 2019 [1] | Three drug-type cultivars, 12/12 | Visible stigmata day 8 (one cultivar) and day 12 (two cultivars); sampling every 5–7 days, so coarse |
| Zhang et al. 2021 [4] | 27 hemp cultivars, 12 h days | Averaged 13–14 days; fastest essential-oil types ~13 days, slowest ~21 days; fibre types from the far south took 21–25 days |
| Peterswald et al. 2023 [6] | Three medicinal lines at 10, 12, 14 h | One line had pistils on 60–73 % of plants by the first check; the other two had few or none then, and all three were at 100 % about 9 days later.being reviewed |
| Ahrens et al. 2024 [7] | Two THC-dominant cultivars, 12 h vs 13 h | 13 h delayed onset by about 1.5 days in one cultivar and not at all in the other (per-treatment day counts are in the paper's figure) |
| Lisson et al. 2000 [8] | Two fibre hemp cultivars, 8–16 h | Rapid flowering below about 14 h; flowering delayed more the longer the day, using thermal time from emergence |
Reading the table. Most cultivars studied showed onset between about one and two weeks after the switch, which matches the house figure of 10–14 days. It isn't universal: some northern hemp flowered in 4–11 days, and slow cultivars took three weeks [4]. Most of the data are hemp or a handful of drug-type cultivars, so treat 10–14 days as the typical case and measure your own (Supported).
The stretch belongs to the transition
The stretch isn't random; it's part of the switch. Under short days, the main stem and internodes elongated fastest from about day 5 to day 10, and short-day plants outgrew long-day plants over the first 14 days, before the condensed inflorescence took over [3]. A US plant-patent description puts the general figure at 50–100 % height gain in the first two to three weeks after floral induction, and 71 ± 13 % for the cultivar described [10]. That's a breeder's description, not a controlled multi-cultivar trial; grower reports run from about 1.5× to 3× (Grower practice). See stretching.
When the stretch slows and the new internodes stay short, the inflorescence is forming. That's a better sign of day 1 of flower than the calendar.
Cultivar variation and the critical photoperiod
Day length decides how fast induction happens, and the threshold varies by cultivar.
- Explants of one drug-type genotype flowered best at 12.0 and 13.2 h; the authors concluded it needed at least 10.8 h of uninterrupted dark [5].
- In hemp, essential-oil cultivars flowered 1–2 days later above 13 h and 7–8 days later above 14 h; some had a critical photoperiod between 13 h 45 min and 14 h, and as little as 15 minutes changed floral initiation in some [4].
- At 14 h every line in Peterswald et al. still flowered and made more flower mass, but two high-THC lines fell from 12 % to 7 % and from 10 % to 6 % THC [6].
- At 13 h, one cultivar flowered about 1.5 days later and the other not at all, with 1.35× and 1.5× the inflorescence yield and cannabinoid concentrations equal or higher. The study ended before full commercial maturity [7].
- Major-effect flowering-time loci (Autoflower1, Early1) have been mapped, so timing is partly simple genetics [9]. Autoflowers skip this question entirely; see photoperiod vs autoflower.
House position: 13/11 is endorsed, and some cultivars flower at 14–15 h (in practice about 12 h at high intensity, plus sunrise and sunset ramps). The data above support it: 13 h barely moved flowering onset in the cultivars tested [7], and every line flowered at 14 h [6]. Expect a longer photoperiod to add a day or two to the transition in sensitive cultivars [4][7], and watch THC concentration at 14 h [6]. Test your cultivar.
What she's telling you
Signs that flowering has really started, in order of how much they can be trusted:
- Pistil clusters at the top nodes and the main cola, not lone preflowers. Three or more pairs of pistils at the top of the main shoot is the research definition of onset [1][7]. A single pair in a leaf axil doesn't count; those form in veg too [1][2].
- The stretch slows and internodes shorten. Short days condense the apex: elongation peaks around days 5–10, then the internodes stay short as bracts stack [3].
- New growth switches from fan leaves to bracts. The apex branches into bract and flower primordia instead of making big leaves and long shoots [1][3].
- Leaf size, aroma and water use change. New leaves at the tops come in smaller with fewer leaflets, the plant starts to smell, and water use shifts as growth moves from stems to flowers (Grower practice; the smaller leaves and fewer leaflets near the inflorescence are consistent with the architecture in [1], the rest is grower observation).
Photo-log tip. Shoot the same top cola from the same angle, at the same time of day, every 2–3 days from the flip. Side by side, the day the internodes stop growing and the pistil clusters appear is obvious. Note that date in your journal as day 1 of flower.
How to count: days since flip vs days of flower
Keep two clocks.
| Clock | Starts | What it's for |
|---|---|---|
| Days since flip | The first long night | Planning the room, light schedules, stretch, comparing with your own past runs |
| Days of flower | First pistil clusters at the top of the main shoot (3+ pairs) | Comparing with breeder flowering times, harvest windows, trichome checks |
Many breeders' stated flowering times appear to count from flower set, not the flip (Grower practice, house position)being reviewed. If you don't know which a breeder means, assume flower set and add the transition.
Worked example: a breeder's "8 weeks"
- Stated flowering time: 8 weeks = 8 × 7 = 56 days of flower.
- Transition from flip to flower set: 10–14 days (house figure; consistent with [1][4]).
- 56 + 10 = 66 days; 56 + 14 = 70 days after the flip.
- 66 ÷ 7 ≈ 9.4 weeks; 70 ÷ 7 = 10 weeks.
- Plan for harvest about 9.5–10 weeks after the flip, then decide on the day by the trichomes, not the calendar (see trichome ripeness).
If you flower at 13/11 or longer, allow another day or two for a sensitive cultivar [4][7].
Common mistakes
- Counting the flip as day 1 of flower, then harvesting two weeks early because "it's week 8".
- Calling preflowers in veg "flowering". They form under long days [1][2].
- Light leaks or a broken timer in the first two weeks. Induction can stall or reverse while it's still being set up [3]. Make the room dark.
- Running out of headroom: most of the stretch comes in this window [3][10]. Flip at about half the final height you can fit.
- Comparing your run with a breeder's number without knowing which clock they used.
Evidence
- Short-day induction and cultivar-specific critical photoperiods: Established [4][5][6][8].
- Visible inflorescence typically appearing about 1–2 weeks after short days begin: Supported [1][4][6]; mostly hemp and a few drug-type cultivars; some cultivars much faster or slower [4].
- Solitary flowers ("preflowers") being age-dependent and photoperiod-independent: Supported [1][2][3].
- Induction needing continuous short days, and long days reversing the inflorescence: Supported by one study [3].
- Stretch concentrated in the first two to three weeks: Supported for timing [3]; size from one patent description [10] and Grower practice.
- 13 h adding little or no delay: Supported by one study of two cultivars [7]; 14 h lowering THC in high-THC lines: Supported by one study [6].
- Breeder flowering times counting from flower set, and runs going long by about the transition: house position (founder's experience) and Grower practice; not measured in any study we've found.
