English 
Most people buy more seed than they plant immediately, which means most people store seed, which means most people are quietly deciding how long it stays viable without thinking about it. Storage is the one part of the whole process that is entirely under your control, costs nothing, and is almost never discussed.
What you are actually protecting
A seed is a dormant living organism with a finite energy reserve and the cellular machinery to use it. Dormancy slows that machinery down enormously but does not stop it. The seed is still, very slowly, running.
Storage is the business of keeping that rate as low as possible. Everything that raises it, whether heat, moisture, light or disturbance, spends reserves that are needed later for germination. A seed stored badly is not damaged in any visible way; it has simply used up a portion of what it needed.
This is why viability declines gradually rather than expiring on a date. There is no cliff, only a slope, and storage decides how steep it is.
The four variables
Temperature
Heat raises metabolic rate directly, so cool is better than warm. More important than the absolute value is stability: every cycle of warming and cooling costs something, and a cupboard that stays moderate beats one that is sometimes cold and sometimes not.
Moisture
Humidity is the quietest threat because it is invisible. Damp conditions encourage the seed to begin processes it cannot complete, and they invite mould. Dry is essential, and an airtight container is the simplest way to achieve it.
Light
Light contributes to degradation independently of heat. Dark storage is trivially easy to arrange and there is no reason not to.
Disturbance
The least obvious of the four. Repeatedly opening a container exposes the contents to room air and room humidity each time, and handling seeds adds pressure and warmth. Opening a packet to look at the contents is not free, and doing it monthly over a year adds up.
A practical setup
The whole thing can be solved with an airtight jar in the most temperature-stable cupboard in the house, and for most people that is the end of it.
If you want to improve on that, add a small desiccant sachet to control humidity inside the container, and keep the container away from exterior walls, windows and anything that generates heat. Label everything at the moment of storage, because identical packets become anonymous within weeks and an unidentified seed is worth considerably less than a labelled one of the same quality.
A refrigerator works and extends viability further, with one important condition: the container must genuinely seal, because condensation forms on a cold container the moment it is brought into a warm room. Taking a sealed jar out and letting it reach room temperature before opening avoids this entirely. A jar opened cold does not.
What matters less than people think
Freezing is frequently recommended and is genuinely effective for long-term storage, but it is unnecessary for the timescales most growers work with and it introduces a failure mode: any moisture present when freezing can damage the seed directly. For a collection that will be used within a few years, cool and stable is sufficient and considerably safer.
Vacuum sealing is similarly more than most people need. Airtight is the useful property; removing the air entirely adds little for a packet that will be opened within a couple of seasons.
And seed appearance tells you nothing about storage condition. A seed that has been kept badly looks exactly like one that has not.
How long is realistic
Well stored seed stays usable for considerably longer than most people assume, with germination rates declining slowly rather than sharply. Established lines such as Northern Lights feminized or White Widow feminized are no different in this respect from recent releases; storage, not genetics, is what determines the curve.
Badly stored seed, by contrast, can lose a significant proportion of its viability within a single warm, humid year. The difference between the two outcomes is a sealed jar and a sensible cupboard, which is a remarkably small intervention for the size of the effect.
If you are returning to a packet that has been stored for a long time, plant more than you otherwise would rather than assuming the worst. Critical 2.0 feminized and lines like it frequently surprise people who had written off an old packet.
Storing an opened packet
A situation that comes up constantly and gets handled badly. You buy five seeds, plant two, and the remaining three go back in a packet that has already been opened once.
That packet is no longer sealed in any meaningful sense, whatever it looks like. The useful move is to transfer the remainder into an airtight container rather than relying on the original packaging, and to do it immediately rather than at some later point when the packet has already spent a month absorbing room humidity.
Keeping several part-used packets inside one larger airtight jar works well and solves the labelling problem at the same time, since the packets themselves carry the names.
Bringing seed home
The period most often overlooked is the first few days after purchase, before anything has been organised. Seed left in a bag by the door, in a car, or on a windowsill for a week has had a worse time than it will have in the following year of proper storage.
Putting a new purchase straight into its final container is the single highest-value minute in the whole process, precisely because the conditions it would otherwise sit in are usually the worst it will encounter.
What to record
Two things are worth writing on or with each packet: the variety and the date you acquired it. Neither is interesting at the time and both become valuable later, when you are deciding how many to plant from a packet whose age you would otherwise be guessing at.
Anyone keeping more than a handful of varieties benefits from a single list rather than relying on the packets themselves, because a list can be read without opening the container.
Why viability declines at all
It is worth understanding the mechanism, because it makes the advice obvious rather than arbitrary. Dormancy slows a seed metabolism dramatically but never stops it, and everything it does while dormant is paid for out of a reserve that cannot be topped up.
Warmth raises the rate of that spending. Moisture raises it further and adds the risk of processes starting that cannot finish. Over months and years the reserve shrinks, and at some point there is not enough left to push a root out and reach light. Nothing dramatic happens at that point; the seed simply fails to start.
Every storage recommendation follows directly from that. Cool, dark, dry and undisturbed all mean the same thing in practice: spend as little of the reserve as possible before you need it.
Testing a doubtful packet
If you have an old packet and want to know where you stand before committing a cycle to it, the sensible approach is to start a small number rather than the whole packet. Three seeds give a usable indication, and the result tells you how many to plant from the remainder.
What is not worth doing is judging by appearance, pressing seeds to test hardness, or any of the folk methods that circulate. None of them correlate reliably with viability, and several of them cause the damage they claim to detect.
Frequently asked questions
Can I store seeds in the freezer?
You can, and it extends viability, but it is unnecessary for most timescales and risks damage if any moisture is present. Cool, dry and stable is sufficient and safer for a collection in regular use.
Does a refrigerator work?
Yes, provided the container is genuinely airtight and you let it reach room temperature before opening. Condensation on a cold container opened in a warm room is the main risk.
How can I tell if old seed is still good?
Appearance tells you very little. The only reliable test is to plant some, which is why planting a few more than usual is the sensible approach with an older packet.
Does the original packaging matter?
Less than where you put it. A sealed container in a stable cupboard protects seed regardless of what it came in; the original packet in a warm drawer does not.
