Rules-of-thumb projections for orthodox seed. Not a test of your seed, and not a germination result. It applies published seed-count tables, Harrington's storage rules and the Ellis-Roberts viability equation to estimate how long dry-stored orthodox seed stays viable. It does not know your seed lot and cannot tell you whether a specific packet will germinate. Only a germination test can. Use at your own risk. Open for the full scope limits.
Model scope
This calculator applies published seed-count tables, Harrington's storage rules of thumb and, for the species that have published constants, the Ellis-Roberts viability equation. It estimates how long orthodox seed stays viable under dry storage. It does not test your seed, does not know your seed lot, and cannot tell you whether a specific packet will germinate.
A projection is not a measurement
- Storage-life figures are projections from thumb-rules, fitted across species and conditions. They describe a tendency, not your packet.
- The only way to know whether seed is still good is a germination test: 20 to 100 seeds on damp paper towel at the species' optimum temperature.
- Starting viability matters more than storage time and is invisible here. Seed harvested immature, dried too fast, or damaged in threshing begins below par, and the model assumes it did not.
- Harrington's rules assume steady conditions. Real storage cycles in temperature and humidity, and cycling is worse than a steady average at the same mean.
- Nothing here models disease, insect damage, rodents, or a container that was not actually airtight.
What the viability equation does not see
- It assumes airtight storage at a constant temperature and a constant seed moisture content. It was fitted to seed sealed in hermetic containers. Seed in a paper packet takes up moisture from the room, and at low moisture content seed in open storage dies much sooner than the equation says.
- It needs the moisture content of the seed. Room humidity is a different quantity. At the same 15% relative humidity lettuce seed holds 4.2% water and barley 6.2%. Entering a guess moves the answer several-fold, and the tool has no way to convert humidity to moisture content for you.
- The constants describe a species, fitted from a few seed lots. Where two published determinations exist they commonly differ two- or three-fold, and the tool shows both.
- Your lot's starting quality comes from one germination figure. A real estimate of it needs an ageing test, and seed weakened by harvest or drying history starts lower than its germination percentage suggests.
- The equation was fitted between -13 and 90 °C. Freezer temperatures are an extrapolation, and nothing below -20 °C is evaluated.
- Drying below a species-specific limit, somewhere between 2 and 6% moisture, gives no further gain. That limit is recorded for only five species here.
What the container model does not see
- The oxygen factor was measured on one species. It comes from seven years of primed celery (Groot et al. 2025). Other work on lettuce found a much smaller effect, so oxygen control only raises the upper end of a projection.
- It cannot tell whether your jar is actually airtight. In the one test that compared closures, plastic screw caps let oxygen back in and rubber rings and lined lids did not, but every jar and lid differs.
- The fall in oxygen inside a sealed jar is scaled from dry lettuce at 20 °C. The rate depends on temperature, moisture and the oxygen level, and no other seed or condition has been measured.
- It does not know what is in your absorber packet. Most contain a moisturizer, which is why an absorber needs a desiccant beside it.
- Sealing seed that is not dry makes it die faster. The tool flags it, but it cannot measure your seed's moisture for you.
Recalcitrant and intermediate species are refused
Refused by design, because dry-storage models do not apply to them. A refusal is an answer, and the right response is to look up the storage behavior of that species rather than to find a tool that will return a number.
Never use this for
- Deciding a seed lot is viable, or that it is not, without testing it.
- Seed bank accession, deaccession or regeneration scheduling.
- Commercial seed sale, labeling, certification or a germination claim, all of which are regulated and require assay data.
- Planting or crop planning where a failed stand is expensive or a season is lost.
- Anything about a species whose storage behavior is not orthodox.
Verify independently
Check the cited sources and run a germination test before relying on any result for a planting, a seed bank, or a commercial decision. A test on the actual lot settles the question; nothing on this page does.
Inputs
Pick a species by common name, or work from seeds you counted yourself.
Settings auto-save to this browser as you type.
How To Use This Tool
This tool answers three separate questions. Mixing them up is the usual source of confusion.
- How many seeds are in this packet? Search the species, read the seeds-per-ounce card. If two sources disagree the tool shows both and does not average them: averaging a transcription error with a correct value produces a number that is wrong and looks authoritative.
- How many seeds do I actually have? Weigh a sample, count it, and enter both under Count. A measured count beats every table here. Published counts carry cultivar, seed-lot and cleaning-standard variation; your own packet has none of that. The published range then becomes a cross-check.
- How long will this seed keep? Set the storage temperature and moisture. The multiplier is relative to the baseline conditions the published longevity figure is assumed to describe, so the baseline is an editable input.
Moisture content and relative humidity are different quantities. Seed moisture content is water as a percentage of seed weight, typically 5 to 14% in stored seed. Relative humidity describes the air around the seed. Harrington's moisture rule uses moisture content; the Hundred Rule uses relative humidity. Both therefore appear as inputs.
Temperature is shown as a range. Harrington's rule is published both as halving per 10 °F and as halving per 5 °C, and the Ellis-Roberts equation's temperature terms, fitted on measured seed survival, give a smaller effect than either below about 35 °C. The projection spans all three, and the card names which one sets each end.
Freezer storage is outside these rules. The tool applies Harrington's thumb-rules between 0 and 50 °C, the range Harrington gave for the temperature rule. Ask for -18 °C and the thumb-rule projection clamps to 0 °C and says so. For the species with published constants, the Ellis-Roberts viability equation answers the freezer question instead, down to -20 °C.
The viability equation answers a different question. Harrington scales a published "keeps about N years" figure. The equation predicts germination percentage against time for a lot that starts at the germination you enter, sealed airtight at the temperature and seed moisture content you enter, and reports how long until it falls to the floor you set. It takes the seed's own moisture content. Room humidity is a different quantity: an oily seed like lettuce holds far less water than a cereal in the same air.
The container matters as much as the fridge. Under Storage, say what the seed is kept in. A sealed jar keeps the seed as dry as it was when sealed, which helps dry seed and harms damp seed. Most oxygen absorbers carry a moisturizer and need a desiccant beside them, and their benefit takes years to show. When you take seed out of cold storage, let the closed container reach room temperature before opening it, or water condenses on the cold seed.
The tool will refuse some species. Acorns, chestnuts, avocado and mango have recalcitrant seed that dies on drying; garlic and potatoes are not grown from stored seed at all. In both cases the tool withholds the storage-life model and explains why. Seed counts still display.
Find a species
Count your own sample
Your packet
Published counts for this species
| Source | Listed as | Seeds / oz | Seeds / lb |
|---|