Canada's Premium Cannabis Seed Bank | Discreet Shipping
Canada's Premium Cannabis Seed Bank | Discreet Shipping
A screen does not make light. It only decides which bud sites get to stand in the light you already paid for. That is the whole idea behind SCROG cannabis growing, and it is worth saying plainly up front, because most of what gets written about screens implies the net itself is doing something magic.
It isn’t. The net is a jig. You are the one moving the plant.
This post is about canopy-level training — the screen, the net, the number of plants you put under a light. The per-plant cuts and bends live elsewhere on this site, with links as we go. The aim here is the honest version: what the research actually supports about SCROG cannabis canopies and plant density, what nobody has tested, and how to run a screen without wrecking your grow in week four.
SCROG is short for Screen of Green. You stretch a net — usually 5 cm (2 inch) squares — horizontally over your pots, and as the plant grows up through it you tuck the shoots sideways under the mesh instead of letting them climb. Branches that would have become a tall centre cola with a stack of shaded runts underneath get spread out flat, each one occupying its own square, each one ending up at roughly the same height.
Why does that matter for a SCROG cannabis grow? Because the difference between the top of a cannabis plant and the bottom of it is much bigger than most growers assume.
Danziger and Bernstein measured it. In a 2022 greenhouse trial published in Frontiers in Plant Science, they sampled inflorescences from five locations on each plant and found that cannabinoid concentrations in the axillary inflorescences at the bottom of the plant were up to 90% lower than in the apical inflorescence at the top. In their untouched control plants, moving from one plant per square metre (about 11 sq ft) to two knocked all six detected cannabinoids at the bottom of the plant down by 71–76%. They attribute the drop to light not getting through the denser canopy, and their light measurements fit that explanation.
One caveat that matters, stated up front rather than buried: their model plant was a type III cultivar — high CBD at 8–16%, THC under 1%. It is not a high-THC drug-type plant. So read that 90% figure as what it is, a light-distribution result, not a potency promise about the strains you’re likely growing.
Still, the shape of the finding is the case for a screen. The lower a bud site sits, the less it is worth. A SCROG cannabis setup doesn’t create new bud sites. It relocates existing ones into the top layer where the light is.
Here is the part the grow guides skip.
There is exactly one peer-reviewed ScrOG grow study we could find, and it has no comparison group. New Zealand’s Crown forensic science institute ran three cycles of six plants under a screen, published in 2010, to work out what a typical indoor hydroponic operation actually produces. It reported an average of 687 g (24 oz) of dry flower per plant across the 18 plants (Knight et al. 2010).
Do not read that as what a screen gets you: it is a forensic yield estimate from a grow deliberately set up to mimic the hydroponic operations commonly encountered by New Zealand Police, the authors call their own numbers conservative because the growers hit problems in all three cycles, and since every plant in the study was screened, nothing in it isolates what the net contributed. Nothing was grown alongside them without one.
So we could not find a controlled trial that grew cannabis with a screen and without one and compared the yields — which is the comparison you’d actually want. The rest of the evidence for SCROG cannabis training is indirect: we know light falls off sharply through a canopy, we know lower buds are chemically poorer, and we know from the same research group’s 2021 work that deliberately reshaping a plant changes how uniform its flower is (Danziger & Bernstein 2021). Stringing those together into “a screen raises yield” is a reasonable inference. It is not a measured one, and any SCROG cannabis yield percentage you’re quoted is unsupported by any controlled comparison.
That is not a reason to skip a SCROG cannabis screen. It is a reason to be honest about what you’re aiming for: a more even canopy, with more of the flower up in the light and less of it as shaded popcorn at the bottom. Nobody has measured how much. If you want the measured evidence on leaf removal specifically, we went through it in does defoliation actually increase cannabis yield.
Sea of Green, or SOG, solves the same problem from the other direction. Instead of training one plant to fill a square metre the way a SCROG cannabis screen does, you put several small plants in that square metre, veg them barely at all, and flip. Each plant gives you one main cola. The flat canopy comes from plant count rather than plant training.
It is genuinely faster. A SOG run can be flipped after one to two weeks of veg, which pulls weeks off the calendar compared to filling a screen. If you’re running back-to-back cycles, that is real money.
Whether SOG also buys more weight per square metre than SCROG cannabis training is where the research gets interesting, because the studies point in different directions.
The same 2022 trial found that doubling density from one to two plants per square metre reduced yield per plant but raised flower weight per area by roughly 30–80% in three of its four treatments (Danziger & Bernstein 2022). Good news for SOG. But cannabinoid yield per cultivation area — the milligrams of active compound coming off each square metre — was not affected by density or by architecture treatment.
And the one treatment that looks most like a SCROG cannabis plant, topped early so it grew on its lower branches, gained nothing from doubling density and, at one plant per square metre, produced as much flower per square metre as any of the four (the paper didn’t test the treatments against each other on that measure).
A 2024 German trial pushed density much further, from 12 to 36 plants per square metre, and found flower yield rose from 119 to 247 g/m² and CBD yield from about 4.3 to 9.4 g/m², with a bigger share of the flower coming from the top of the canopy (Schober et al. 2024). That was one CBD cultivar, vegged in a greenhouse and flowered under lamps at fairly low light (roughly 150–450 µmol/m²/s), at densities far above any home tent.
The same group also found each extra week of veg, from one to four, added roughly 100 g/m², which is the price of Sea of Green’s short veg. And unlike the 2022 trial, crowding didn’t significantly steepen the top-to-bottom potency gradient.
Then there’s the wider literature. Backer and colleagues pooled what published cannabis yield data they could use in a 2019 meta-analysis in Frontiers in Plant Science — only three studies cleared their bar, because nothing else reported growing conditions in enough detail — and found that increasing plant density reduced yield per watt and THC per square metre, and that plant density “was not an effective predictor of yield per square meter.” Their conclusion recommends low density — 12 plants per square metre or fewer — for yield per square metre and per watt.
The authors are careful about one thing, and so should we be: in their dataset, density was so tightly correlated with maximum growing temperature and with the length of the vegetative photoperiod that the three effects cannot be separated. So the density research is mixed. More plants can mean more weight per square metre, but not reliably more cannabinoids, and the density lever is not the free money it looks like.
Worth noting that a home-scale SOG of four or five plants per square metre is still comfortably inside the low-density band that meta-analysis recommends — the number it lands on is 12 per square metre or fewer, so four in a tent is not what it is warning about.
The honest read, and it is a read rather than a finding: SOG buys you speed and simplicity. SCROG cannabis training buys you an even canopy off fewer plants. Neither one conjures grams out of nothing.
| SCROG | Sea of Green | |
|---|---|---|
| Plants per m² (about 11 sq ft) | 1–2 | 4 or more |
| Veg length | 4–8 weeks | 1–2 weeks |
| Seeds needed | Few | Many |
| Best seed type | Stretchy photoperiod | Compact photoperiod or auto |
| Main cost | Your time, weekly | Seed spend, every cycle |
| Fails when | You flip too late | Airflow dies in the crowd |
Every row in that SCROG cannabis comparison is grower practice, not a trial result.
Legal plant limits usually settle this argument before horticulture gets a say. Canada allows four plants per household, with Quebec the one province that still bans home growing outright; US limits vary by state. If you’re capped at four plants, a SOG can only fill about a square metre (11 sq ft) at its usual spacing. In anything bigger, a SCROG cannabis screen is how four plants fill the space. We broke the plant-count maths down in how many cannabis seeds should you buy.
Too low and you’re weaving seedlings that have no side branches yet. Too high and the plant builds a tall bare stem under the net that does nothing for you. Twenty to thirty centimetres above the top of the pot is the range most growers settle on, and it gives you room to get a hand underneath.
This is the rule people break, and the advice you’ll find is all over the place — plenty of guides say wait until the net is nearly full, and none of it is research-backed either way. Err low.
Here is the reasoning. Growers commonly report a balanced hybrid gaining roughly 1.5–2× its height after the flip and a sativa-leaning plant 2–3×, though the research is split (one greenhouse study, Danziger & Bernstein 2021, saw no post-flip stretch at all), and that growth goes into filling the rest of your net.
Flip a full SCROG cannabis screen and that new growth has nowhere to go but up — there are no empty squares left to tuck it into — and by the end of week three Rule 4 closes the window for good. Wait until the screen is packed before flipping and you’ll spend week three fighting a canopy sitting 30 cm (12 inches) above the mesh. We covered the timing and the size of the jump in the cannabis flowering stretch.
Take each growing tip and tuck it under the mesh into the next empty square, so the plant is held down by the net rather than by string. Growers typically weave every two or three days during veg — soft green growth bends without complaint, woody stems snap. This is low stress training applied to a whole canopy, and it is most of the week-to-week work in a SCROG cannabis grow, and the technique itself is covered properly in low stress training cannabis. The free green 90° clips that ship with every Lighthouse order work well for the first few bends before the screen takes over.
Once the stretch finishes — usually around week three of flower — the shoots are setting flower and the stems are stiffening. Keep pushing them around and you’re breaking branches for no gain. Let the canopy do what it’s going to do.
Anything below the net in a SCROG cannabis canopy is in permanent shade and will produce larf. Strip it. In the 2022 density trial, every treatment except one lost cannabinoid concentration in its low inflorescences when density doubled. The exception was the treatment that stripped leaves, branches and flowers off the bottom of the plant: those low inflorescences came out 25–90% higher at the crowded density than at the sparse one. That supports the quality half of this rule only: the same paper found no change in cannabinoid yield per square metre, so it’s about what the flower you keep is worth, not total output. Cleaning out the basement isn’t just tidiness.
Growers find one vigorous photoperiod plant will typically fill a 1 m × 1 m (about 3.3 ft × 3.3 ft) SCROG cannabis screen with six to eight weeks of veg. Two will do it faster. Four will do it faster still and then fight each other for airflow, which is how bud rot gets started in a dense canopy — see cannabis bud rot for what that looks like before it’s too late.
An autoflower runs on a fixed clock. It will start flowering whether or not your screen is full, so you can’t extend veg to finish the job, and a compact auto was never going to fill a metre of mesh. No trial has tested a SCROG cannabis screen on autos either way. If you want to try, pick a stretchier auto and a small screen, and start weaving early; for most autos, gentle bending while the plant is young is the easier route. The height ranges and why they’re fixed are in autoflower plant height.
Seed choice decides half of this before you unroll a net.
A SCROG cannabis screen wants a plant that stretches and branches — the kind that would otherwise give you one enormous cola and a lot of shaded understory.
Durban Poison is a pure sativa that often doubles or triples in size in flower, up to 400 cm (13 feet) outdoors; it’s rated Easy, and for a few plants per square metre its listing recommends a screen of green. Super Skunk Silver Haze can reach 450 cm (15 feet) outdoors, and its breeder, calling its growth extremely vigorous, advises flipping to flower after just two weeks of veg and running a screen of green indoors. Mac’s own Neptune’s Wedding runs 90–130 cm (35–51 inches), and its page suggests a screen of green and stakes to hold its colas up late in flower.
For something easier to live with, Gorilla Glue #4 runs up to about 137 cm (54 inches), and its listing says a little topping or a screen keeps the canopy even. White Amnesia Haze, at a guide height of 100–200 cm (39–79 inches), is one to train early indoors, and Train Wreck, at 100–150 cm (39–59 inches) indoors, is listed as responding well to topping.
Sea of Green wants the opposite plant: short, fast, one dominant cola, minimal side growth to waste energy on. White Widow Autoflower at 70–80 cm (28–31 inches) is about as compact as we carry, and Mandarin Cookies Autoflower is expected to stay at up to about 100 cm (39 inches) and takes about 8–9 weeks from seed as a rough guide. If you’d rather keep photoperiod control and still run tight, Bubba Kush tops out around 100 cm (39 inches) indoors, with very few side branches. Blue Dream is tall, listed at 300–400 cm (9.8–13.1 ft) outdoors, but Leafly notes it suits a sea of green, with a short vegetative period recommended indoors.
Browse the full feminized cannabis seeds and autoflower cannabis seeds ranges if you want to filter by height yourself — the listed plant height on every product page is the number that matters here.
If you’re limited to a handful of plants — which most legal home growers are — the screen is the better tool for anything bigger than about a square metre. SOG’s advantage is speed, and a plant cap limits how much floor it can cover.
But keep the claim modest. A SCROG cannabis canopy does not manufacture yield. It moves bud sites up into the light you already bought, away from the shaded bottom of the plant where, in the research, flower carries far lower cannabinoid levels. On that narrower point the research is on your side, and that is enough reason to string a net.
Start with a plant that wants to stretch, hang the screen low, and flip earlier than feels right.
Shop Feminized Seeds | Low Stress Training Cannabis | Maximizing Yield in Cannabis | More cannabis growing guides
Probably, but no controlled cannabis trial we could find has measured it. What is measured is the light gradient a screen is meant to fix: bottom inflorescences carrying dramatically lower cannabinoid concentrations than the top of the same plant. A SCROG cannabis canopy attacks that gradient. The aim is more even flower and less popcorn, and treat any specific percentage you read as unsupported.
One or two per square metre (about 11 sq ft) for a SCROG cannabis grow. A single healthy photoperiod plant will typically fill a 1 m × 1 m (about 3.3 ft × 3.3 ft) net given six to eight weeks of veg. Adding plants shortens veg but crowds the canopy. The density research is mixed: one trial found doubling from one to two plants per square metre raised flower weight per square metre but not cannabinoid yield, another found both rose between 12 and 36 plants per square metre, and a three-study meta-analysis found density didn’t predict yield per square metre.
You can try, but a SCROG cannabis screen is hard to do well on an auto, and no trial has tested it. The fixed clock means you cannot hold veg long enough to fill a big screen, so it suits a stretchier auto under a small screen at best. For most autos, gentle bending while the plant is young is the easier route.
Around 5 cm (2 inch) squares is the grower convention for a SCROG cannabis net. Nothing peer-reviewed sets this number — it’s practice, not research.
No. Most people do both: top once or twice in early veg to get multiple mains, then use the screen to spread those mains out. The two techniques answer different questions, and the cutting side is covered in topping cannabis vs FIMming.
Keep reading
Cannabis Yield
Does Defoliation Actually Increase Cannabis Yield? What a 2021 Study Found
Maximizing THC PotencyBrowse all guides to growing cannabis by topic, from germination to harvest.
Ready to grow? Browse premium seeds from Lighthouse Genetics:
Designed & Built by Wynne Web Design
