Can trichome colour tell me when to harvest?
Less than the loupe ritual implies — and the reference text for the field says so plainly.
By Curtis Olson — retired software developer, home grower, and the builder of the device this site is about. More about me and why I built it. · Last reviewed
The short answer
Trichome colour is a maturity cue, and it’s the best cheap one available. What it is not is a readout of what’s in the plant. The link between the colour you see through the loupe and the actual cannabinoid and terpene composition is, in the words of the standard reference on controlled-environment cannabis production, still poorly understood.
Everyone in this hobby, myself included, has been taught to read cloudy-versus-amber as if it were a dial showing chemistry. It isn’t one.
“Milky” and “cloudy” are the same thing, so if you came here searching for one, this page is about the other. Clear heads turn milky-white before they turn amber, and which word a grower reaches for is habit rather than a distinction — the loupe shows one transition, not two. The words that follow use cloudy, and none of the argument changes if yours is milky: the problem is not which word names the colour, it is that the colour is being read as a readout of chemistry it cannot give you.
Why “it looks ready” is not an answer — the three problems
They don’t mature together. Trichomes on a single plant — and even on a single bud — are at different stages at any moment. So whatever you saw at 30× is a sample of one small patch, and which patch you chose changes the answer. This is exactly why two experienced growers can look at the same plant and disagree by a week.
The colour-to-chemistry link is unresolved. The practice is near-universal; the evidence connecting the visual stage to the composition of what you end up smoking has not caught up with it. That’s not me being contrarian — it’s the position taken in the Handbook of Cannabis Production in Controlled Environments, which is about as mainstream a source as this field has.
The colour you see is partly your lamp’s. This is the one nobody warns you about, and it sits underneath the other two: before you get to what the colour means, there is a question about whether you are seeing the colour at all. It comes up constantly in help threads, in three flavours that are really one problem. The loupe’s own LED washes the heads and makes clear ones read as cloudy. A bead of water or leftover spray refracts and throws an amber glint that isn’t pigment. And under a blurple lamp you cannot judge any colour reliably — growers routinely find that a leaf they were sure was deep green turns out to be yellowing the moment they carry it out into white light.
None of that is carelessness. It is optics, and it applies to everyone equally, which is why two people at the same bench with the same loupe can still disagree.
“How much longer?” “Is it time?” “Is it ready?” “Ready to chop?” — what to use instead of the colour
Those are the two ways this gets asked, and they get the same answer. Keep using the loupe — just demote it from oracle to one input among several, and give it company.
The most useful hard number I’ve found: modern cultivars generally mature 45 to 70 days after flowering is induced. Cultivars that run past 70 days are considered less economically worthwhile, because the extra weeks don’t pay for themselves in annual yield.
That range is more useful to a hobbyist than it first sounds, because it’s independent of your eyes and independent of the seed packet. If you’re at day 38 and convinced by the trichomes that it’s ready, one of those two things is wrong. If you’re at day 75 and still waiting for more amber, you’re probably past the point of getting anything back for it.
Combine: days since the flip, the breeder’s stated window, the state of the pistils, and the trichomes. When they agree, cut. When they disagree, the disagreement is the information — go and find out why.
One warning about combining them, because this one trips people up constantly: there are two different “70 %” rules in circulation, and they are not about the same thing. “Wait for 70 % of the pistils to brown and curl” is a claim about the hairs. “Go for 70 % amber” is a claim about the trichome heads. They get quoted interchangeably, the number does the talking, and nobody says which one they mean — so two people can agree on “70 %” while describing harvests a fortnight apart. I watched exactly this happen in a help thread: a first-time outdoor grower said he’d been told to wait for 70 % of the pistils, and the first experienced grower to answer him replied that he always goes for 70 % amber. Neither of them noticed they had swapped metrics. (That’s something I read on a forum, not a citation.)
So when someone hands you a percentage, the useful question is “seventy per cent of what?” — and then treat the answer with the scepticism the rest of this page argues for, because an amber percentage is the least reliable of the four signals above, not the most.
And when you do look, take the light seriously, because it costs nothing to control. Carry the sample out from under the grow light and look at it in ordinary white light. Let the surface dry first — a wet trichome is a lens. Look at the same spot each time rather than hunting for the patch that agrees with you. Judging the pattern across the bud beats judging an absolute colour, and it is the judgement your eyes are actually good at.
Look at the bud, not the sugar leaf
Where you point the loupe is the first problem on this page turned into a habit. The standing advice among growers is to read the heads on the flower itself, the swollen calyxes, and not on the small sugar leaves that stick out of it. The leaves are widely reported to go amber first, so a sample taken from them makes the plant look further along than the flower is. That is practice, not a published measurement, but it is the same point as above: which patch you choose changes the answer, so choose the patch you are going to smoke.
It is also why one alarming photo is not a verdict. In a help thread, one grower described nearly cutting two weeks early because a picture showed a lot of amber, then looking again the next day and finding mostly clear heads. Another, told to stop reading the sugar leaves, said that checking early had made him panic. (Those are things growers wrote on a forum, not citations.) Look at the flower, look in more than one place, and look again tomorrow before you believe today.
I already chopped early. What did I lose?
Less than you have been led to fear, and less than anyone can honestly put a number on. This one turns up constantly, usually a day or two after the fact, and it is almost always asked by someone who has already decided they ruined it. Start with the arithmetic instead of the regret: count the days from the flip. If you landed inside the 45 to 70 day range above, you cut inside the window the reference text gives for modern cultivars, and the likeliest answer is that you are fine.
The uncertainty runs both ways, and here it runs in your favour. The whole argument of this page is that the link between trichome colour and what is actually in the plant is poorly understood. Nobody ever applies that in this direction: if cloudy-versus-amber is not a reliable dial going up, it is not a reliable dial going down either. Anyone who tells you that cutting at mostly-cloudy cost you some specific fraction of anything is claiming a sharpness the field does not have. I can’t tell you what you lost. Neither can they — the difference is that I’ll say so.
What you can say you lost is weight. That part isn’t mysterious. The last stretch of flower is when the bulking happens, so an early cut costs yield in a way you could have put on a scale. If you want one regret to keep, keep that one: it’s the only one you can actually substantiate, and it’s a smaller, duller loss than the one people torture themselves with.
The cure won’t finish the job, and it’s worth being exact about why. Cure properly — curing is the last process that makes the flower better, and the same reference text puts peak quality immediately after the cure. But read that placement carefully. The peak sits immediately after curing, and everything past it is degradation. There is no stage at which cut material picks up where the living plant left off. Curing finishes what you cut; it does not continue what the plant was doing. So a good cure deserves your full attention, and it will not convert an early harvest into a late one.
The fix for next time costs nothing and isn’t a purchase: write down the date you flipped to 12/12. Nearly everyone asking this question can’t answer it, which is exactly why they end up interrogating a photograph instead. With that one date, the question stops being a test of nerve and becomes arithmetic against a published window.
Where this is wrong: if you cut well before day 45 — especially if the pistils were still overwhelmingly white and the buds were visibly still swelling — then you were genuinely early, and the yield cost is real rather than theoretical. And if you’re running one of the long-flowering cultivars that the 45 to 70 day range explicitly doesn’t cover, that window is the wrong instrument for you and the breeder’s own number is the better prior. The argument above is about the ordinary case, which is most cases.
Why I’m telling you this against my own interest
“Point the camera at the bud, get told when to harvest” is the single most requested feature in this category. It is the obvious thing to build, and it would sell.
I’m not shipping it. My device will show you a sharp, magnified, dated view of your trichomes — getting a usable macro image on a schedule was hard enough and it’s genuinely useful, because it lets you compare a week ago to now without having disturbed the plant. What it will not do is convert that picture into a maturity verdict, because the conversion step is the part the literature says nobody has established.
A device that told you “harvest Thursday” would be inventing the confident half. I’d rather hand you a better look at your own plant and let you make the call.
The same goes for the lamp problem above, and this is the part I would rather not have to write. My camera has that disease too, and I can put numbers on it because I went and measured mine. There is no white-balance control in the capture path that ships today. Over one week last summer the colour balance of my own bench pod drifted noticeably toward blue on its own, with nothing about the plant changing. Worse, the auto-exposure loop is deliberately built to hold the rendered green of the leaf near a fixed target — which is to say it is designed to normalise away the exact thing you would want to compare. The frames do not even keep their camera metadata.
So I am not going to offer you a “compare the colour to last week” feature, and I would be sceptical of anyone who does it from an uncalibrated camera. What survives that limitation is pattern — where something is on the plant, and whether it is still changing — and pattern is what I build on. A sharp, dated, magnified look at the same spot over time is worth a great deal. A colour reading from it is worth less than it appears, mine included. This is my own measurement of my own hardware rather than a result from a paper, and I would rather you had it than not.
Sources
- Caplan, D., Matzneller, P. & Gutierrez, G. Harvest and Post-Harvest, in Handbook of Cannabis Production in Controlled Environments (CRC Press). DOI 10.1201/9781003150442-10, §10.2.3 (trichome maturity), §10.3.1 (cultivar timing) and §10.9 (post-harvest degradation — the chapter’s position that quality is at its highest immediately after curing, which is what “the cure won’t finish the job” rests on).
- Mahlberg, P. & Kim, E. (2004) — capitate-stalked trichome distribution.
- Adams, R. et al. (2001) — light intensity and maturation rate. (This one is about light changing how trichomes mature. The third problem above is a separate matter — light changing how they look — and I am not citing this paper for it.)
- The illumination limits in “the third problem,” and the drift figures for my own camera, are first-party observation and my own bench measurements from 2026 — not results from a paper. Labelled that way deliberately, so you can weigh them accordingly.
More of these at Answers. The device is at croplock.com.
Also read: It’s getting cold — do I chop? · Bud rot: how humid, for how long? · Should I flush before harvest? · How do I know my flower is dry enough? · Will the buds still get bigger? · Storing flower after the cure