Is That Old Seed Packet Still Worth Sowing? How to Test Viability Before You Waste a Sowing Slot

You open the seed drawer in February and find it: a half-used packet of parsnip seed from two seasons ago, a few tomato packets with a sell-by date that's long gone, some onion seed you're fairly sure you bought "sometime after the pandemic." The packet looks fine. The seed inside looks fine. But looks tell you almost nothing about whether it will actually grow.

Sow it and hope, and you won't know you've got a problem until three or four weeks later, when the bed is still bare and everything else in the garden has moved on without it. That's a bigger setback here than it sounds: Scotland's frost-free window is shorter than most of the UK's (see the frost dates guide for how much that varies by region), so a failed early sowing often means there isn't enough season left to start again and still get a worthwhile crop, the space, the slot, and sometimes the variety are all gone for the year. Testing a pinch of seed first costs a fortnight and a damp paper towel. Sowing blind and finding out later can cost the whole slot.

The test and figures below are for shop-bought seed, a packet from a garden centre or catalogue that's been sitting since a previous year. Seed collected and dried from your own plants keeps differently, and has its own pitfalls before you even get to testing it (isolation, cross-pollinationPollen moving from one plant to a different, compatible variety, needed for fruit to set on any variety that isn't self-fertile. Full definition →, drying), the saving-your-own-seed guide covers that process and its own viability test.

The paper towel germination test

This is the standard way to get a real answer, rather than a guess, about whether a packet is worth sowing.

1. Count out 10 seeds from the packet. Ten is enough to give you a workable percentage without wasting seed you might still need.
2. Dampen a paper towel so it's moist throughout but not dripping, squeeze out any excess water first.
3. Space the seeds out on one half of the towel, then fold the other half over them (or lay a second damp sheet on top).
4. Seal the towel inside a plastic bag or between a plate and clingfilm, and label it with the crop and the date, it's easy to lose track of which bag is which once you're testing several packets at once.
5. Keep it somewhere warm, an airing cupboard, on top of the fridge, or near (not on) a radiator. Most vegetable seed germinatesA seed sprouting: breaking dormancy and putting out its first root and shoot, given the right warmth and moisture. Full definition → fastest in the high teens to low twenties °C.
6. Check every day or two. Keep the towel damp, mist it or add a little water if it's drying out, but don't let it get waterlogged, since seed sitting in standing water can rot before it gets the chance to germinate.
7. Give it a full two weeks before writing anything off. Gardeners' World's version of this test allows two weeks for germination to show, and some seed, especially anything on the slower side, like parsnip or celery, can take the best part of that to get going even when perfectly healthy. Don't judge the result at day four.
8. Count what's germinated, a visible root (radicle) emerging is enough to count a seed as successful, you don't need to wait for leaves. Multiply the number of germinated seeds by 10 to get your percentage. Five out of ten germinated is 50% viability.

Test each variety separately, even if you're testing several packets in one session, germination speed and rate both vary enough between crops (and sometimes between varieties) that mixing them in one bag makes the count harder to trust.

What to do with the seeds that do sprout: they're proven viable, so it's worth gently potting them on into modulesA tray divided into small individual cells, each sown with its own seed or seedling, so each plant grows undisturbed roots until it's ready to plant out. Full definition → or small pots rather than discarding them, handle the emerging root carefully, as it's fragile and easily snapped. Ten seeds isn't enough of a sowing on its own, though: once you know the packet's real germination rate, sow the rest of it at the thicker or normal rate the result calls for (see below).

Pelleted or coated seed, common for carrot, and increasingly for onion and beetroot, can be tested the same way; just note which packets are pelleted, since the coating makes the seed inside harder to inspect by eye if you're also checking for the dead-seed signs below.

How long does shop-bought seed actually last?

This is where a lot of guesswork happens, and it's worth being specific, because the spread between species is much bigger than most people expect. Some vegetable seed is genuinely a use-it-or-lose-it purchase; other seed will still be worth testing years after the packet date.

The pattern worth remembering: parsnips and alliums are the ones to be suspicious of first. If you find old onion or parsnip seed in the drawer, don't assume it's fine just because everything else in there has kept for years, test it, or budget to buy fresh.

What "stored well" actually means

Seed longevity figures like the ones above only hold if the seed has actually been kept properly in the meantime. A packet that's spent two summers on a sunny windowsill or in a shed that swings from freezing to warm and back isn't the same as one kept consistently cool.

Garden Organic's guidance puts the two damaging factors plainly: moisture and warmth are what shorten a seed's life, and as a rough rule of thumb rather than a precise formula, storage life roughly doubles for every 5°C drop in temperature and every 1% drop in the seed's moisture content. RHS advice on seed storage recommends keeping seed in a refrigerator at around 5°C (41°F), in labelled paper packets or envelopes inside an airtight container with a desiccant (a moisture-absorbing packet or sachet, such as silica gel) to keep moisture out, under those conditions, seed can remain viable for many years.

In practice, "stored well" means:

• Cool, a cold cupboard, unheated room, or the fridge (specifically not the freezer for most vegetable seed unless it's been properly dried first).
• Dark, light and, more importantly, the heat that comes with a sunny spot both work against the seed.
• Dry, damp packets or high humidity are the single biggest cause of seed dying in storage; RHS advice notes that excess humidity or warmth can cause seed to deteriorate or die from fungal disease or rotting.
• Away from temperature swings, a shed, a garage, or a windowsill that heats up in the day and cools sharply at night is worse for seed than a steady, moderate cupboard, even if the cupboard isn't as cold as a fridge.

If your old packet has spent its life in anything other than these conditions, treat the "typical viability" figures above as optimistic best cases, not guarantees, that's exactly the seed worth testing rather than trusting.

Signs the seed is already dead: don't bother testing

Not every old packet needs the full two-week test. A couple of checks take seconds and can save you the trouble:

• Visible mould or fungal growth, either on the seed itself or inside the packet. This is a direct sign the seed has been exposed to damp at some point, RHS advice on seed storage notes that excess humidity or warmth causes seed to deteriorate or die from fungal disease or rotting, and once that's visibly happened, the seed underneath is very unlikely to be worth testing.
• The float test, for medium-to-large seeds (beans, peas, squash, sweetcorn and similar, this doesn't work reliably on very small or lightweight seed like carrot or lettuce). Drop a sample in a jar of water: RHS guidance on seed saving notes that viable seed will sink, while dead seed floats to the surface. It's not infallible, but for larger seed it's a genuinely useful first pass before committing two weeks to a full germination test.

If either of these shows up, skip the paper towel test and go straight to buying fresh.

Reading the result: sow as normal, sow extra, or bin it

Once you've got a percentage from the test, the decision is straightforward, it just needs translating into an actual sowing plan.

These bands are guidance rather than an exact formula, and a crop that isn't your main harvest might be worth sowing at a lower rate than one you're relying on. But they turn "I think this seed is probably fine" into an actual number to act on.

A note for tomatoes, peppers and other seed that needs extra warmth

The table above covers general viability, but tomatoes, peppers and other seed that germinates best above 20°C need the test itself to run warmer than an ordinary cupboard reliably manages, a spot on top of the fridge, near (not on) a radiator, or a heated propagator lid is worth using for these specifically. If the test seed doesn't germinate at all, check the warmth before writing off the packet: a genuinely cold test environment can produce a false "dead seed" result for these crops in particular.

It's also worth separating a dead-seed result from cold, wet soil doing the same thing outdoors later: a packet that tests well can still sit and rot, or simply not come up, if it's sown into ground that's still too cold, the test only rules out the seed, not the conditions it goes into afterwards.

Regional variation within Scotland

A Highland or Aberdeenshire frost-free window runs shorter again than a milder coastal or Central Belt one, so a "moderate, sow-extra" result is a safer bet to act on in a garden with more spare season than one with less. The frost dates guide gives the actual figures by region, check it before deciding how much risk a lower-band result is worth taking on where you garden specifically.

If it's your own saved seed, not a shop packet

Everything above assumes seed in its original packet from a shop or catalogue. Seed collected and dried from your own plants needs the same paper towel test, but there's more to get right before that point, drying, avoiding cross-pollination between varieties, and storage from the point of collection. See the saving-your-own-seed guide for that in full.

Common questions

Sources