How do I test old seed before sowing a whole bed?

Everybody has a tin of half-used seed packets and no idea which are still alive. A failed row costs you the bed, the season slot and three weeks of waiting before you even know it has failed — which is far more expensive than the ten minutes this test takes.

Step by step

1. Count out exactly ten seeds.Ten, so the number that germinate converts straight to a percentage with no arithmetic. Take them at random from the packet rather than picking the biggest.
2. Lay them on damp kitchen towel.Damp, not sopping. Space them out so you can see each one and so a mouldy seed does not affect its neighbours.
3. Fold it over and put it in a labelled bag or tub.A food bag or a lidded box stops it drying out. Write the variety and the date on it — you will be testing several at once and they look identical.
4. Keep it warm indoors.Normal room temperature is right for most crops. Not on a radiator, which cooks them, and not on a cold windowsill, which stalls them.
5. Check every couple of days.Look for a root emerging. Re-damp the towel if it is drying. Fifteen seconds each time.
6. Judge it against the crop's normal germination period.A bean shows in three or four days; a parsnip can take three weeks. Do not write a packet off after a week without checking what that crop normally takes.
7. Count the seeds that produced a root and act on the number.Eight to ten: sow as normal. Five to seven: sow at roughly double your usual rate. Three or four: only worth using where you can afford to thin heavily, not for a whole row. Two or fewer: bin it and buy fresh.
8. Pot up anything that germinated.The seedlings from the test are perfectly good. Lift them carefully with tweezers and pot them on rather than throwing them away.
9. Test the short-lived seeds first.Parsnip, parsley, onion and leek lose viability fastest and are the ones most worth checking. Beans, brassicas, cucurbits and tomatoes are long-lived and usually fine for years.

Tips

• Store seed cool, dark, dry and airtight and it lasts far longer. A sealed box in an unheated room beats a shed or a greenhouse by years.
• The packet's 'sow by' date is a legal minimum, not an expiry. Well-stored seed frequently germinates fine years past it — which is exactly what the test tells you.
• Germination rate is not the whole story: old seed that germinates may also come up slower and weaker, so sow a little thicker than the percentage suggests.
• Doing all your tests in the same week in late winter is far more efficient than testing packets one at a time as you need them.
• Silica gel sachets from packaging, kept in the seed tin, genuinely extend storage life.

Common mistakes

• Abandoning the test after a week — Slow crops like parsnip and parsley take three weeks. Many perfectly good packets are binned days before they would have shown.
• Testing on a radiator — It cooks the seed and dries the towel. Ordinary room temperature is what you want.
• Using twenty or fifty seeds — It works, but you have to do the arithmetic and you waste more seed. Ten is the number that makes the answer instant.
• Sowing marginal seed at normal rate — You get a gappy row and no way to fill it in time. Either sow much thicker or buy fresh.

Questions people ask