Almost everyone who has spent time around cannabis knows someone with a homemade edible story, and it is rarely a good one. That is not carelessness — it is the predictable result of four separate sources of uncertainty stacking on top of each other.
Homemade cannabis edibles safety comes down to understanding where that uncertainty comes from.
Not medical advice. TruMo Analytics is a cannabis testing laboratory. This explains why homemade edible potency is unpredictable. Any question about whether cannabis is appropriate for you belongs with a clinician.
Where the Uncertainty Comes From
1. You do not know your starting material
If your input is dispensary flower with a certificate of analysis, you at least know the potency of what you started with. If it is home-grown, trim, or material of unknown origin, you are guessing at step one — and every calculation after that inherits the error.
2. Decarboxylation is incomplete and variable
Raw cannabis contains mostly THCA, which is not intoxicating. Converting it to THC requires heat, and how completely that happens depends on temperature, time and how evenly the material was heated.
Home ovens have significant temperature variation, and thermostats are frequently off by a meaningful margin. Two people following the same instructions in different ovens get different conversion rates. See cannabinoids explained.
3. Infusion efficiency varies enormously
Cannabinoids are fat-soluble, so they are extracted into butter or oil. How much transfers depends on the fat used, the temperature, the duration, the plant-to-fat ratio and the technique.
Published infusion efficiencies vary widely, and a home cook has no way to know where in that range their batch landed.
4. Distribution is uneven
This is the one that causes the worst outcomes.
Even if the total cannabinoid content of a batch were known, it is not distributed evenly unless the mixture was thoroughly homogenised. In a tray of brownies, one piece can contain several times what another does.
Commercial producers address this with industrial mixing and verify it with homogeneity testing across multiple units from a run. A mixing bowl and a spatula do not achieve the same thing. This is precisely why edibles are tested differently from flower.
Why the Delay Makes It Worse
Ingested cannabis takes 30 minutes to two hours to take effect, sometimes longer, and lasts four to eight hours or more.
That delay is the mechanism behind most unpleasant experiences. Nothing happens, so a second dose goes in, and then both arrive together. With an unknown starting dose and uneven distribution, the result can be several times what was intended.
Ingestion also involves different metabolism — THC is converted in the liver to 11-hydroxy-THC, which is more potent and longer-lasting. An edible dose does not feel like a slower version of an inhaled dose; it feels different. See forms of medical cannabis.
Reducing the Risk
Start with material you have a COA for. Known input potency removes the largest single unknown.
Use an oven thermometer. Not the dial. Decarboxylation is temperature-sensitive and oven thermostats are frequently wrong.
Mix thoroughly and for longer than feels necessary. Uneven distribution is what turns a moderate batch into a dangerous one.
Make small, uniform portions. Individually portioned items — a mould rather than a tray — distribute more evenly than something you cut afterwards.
Assume you overshot. Estimate the dose, then start with a fraction of what you calculated. Wait two hours before any more. This single rule prevents most bad outcomes.
Test a small amount first, on a day when it does not matter. Then treat that as your reference for the batch — but only for that batch.
Write down what you did. Material, quantity, decarb temperature and time, infusion method, yield, portion count. Without a record you cannot reproduce a good result or diagnose a bad one.
Storage and Labelling
This is where homemade edibles cause harm to people who did not choose to consume them.
Commercial edibles carry child-resistant packaging and clear labelling because both are required. Homemade ones look exactly like food, because they are food.
- Label everything — what it is, estimated strength, date.
- Store separately from ordinary food, in a container that does not look like a snack.
- Keep it locked or genuinely out of reach of children. Paediatric cannabis exposures rose sharply as edibles became common, and the majority involve products that were accessible at home.
- Keep it away from pets. THC is considerably more dangerous for dogs than for people, and chocolate compounds the risk. See cannabis and pets.
- Refrigerate perishable items. An infused food is still food, with the same spoilage rules.
Food Safety Still Applies
An infused product is subject to everything that applies to the underlying food. Dairy-based items spoil. Anything containing eggs has the usual considerations. Infused oils stored improperly go rancid.
Cannabinoids also degrade — THC oxidises to CBN with heat, light and time — so a batch stored for months is not the same as it was when made.
When Someone Takes Too Much
It is unpleasant and, in adults, generally not dangerous. What usually helps: a quiet place, reassurance, water, and time. Effects peak and pass.
Seek medical help for severe symptoms — chest pain, difficulty breathing, persistent vomiting, loss of consciousness — and always for a child, an older adult, or anyone with a cardiac condition. Be straightforward with medical staff about what was taken. They need the information to treat, and withholding it delays care.
The Honest Comparison
Commercial edibles are tested for potency per serving and per package, checked for homogeneity across the batch, screened for contaminants, and labelled accordingly. Homemade edibles are none of those things.
That is not an argument against making them. It is an argument for treating an unknown dose as an unknown dose — starting very low, waiting properly, and labelling and storing them as though someone else might find them. Because someone might.
See dosage and administration basics and the risks of untested cannabis products.