Seed Germination Rate Calculator

Introduction to germination percentage, purity and pure live seed

A seed label carries two numbers that look independent and are not. The first is germination, which the USDA Forest Service seed testing manual defines precisely as "the number of normal seedlings produced from 100 pure seeds expressed as a percentage." The second is purity, the percentage by weight of the bag that is actually pure seed of the named species rather than chaff, broken seed, other crop seed, weed seed or inert matter. Germination is measured only on the pure seed fraction that the purity analyst separated out, so germination alone never tells you how much living crop seed is in a bag. Multiply the two and you get pure live seed (PLS), the single number that actually drives a seeding decision.

This calculator is built around that chain. You enter what a seed label or a laboratory report tells you — germination, hard or dormant seed, and purity — plus what your own field records tell you about post-germination survival. It returns the pure live seed percentage on both a label basis and a prompt-establishment basis, the number of seeds you must sow to hit a target plant stand, the bulk pounds per acre that correspond to a recommended pure live seed rate, an empty-cell analysis for plug trays and hills, and a plot-capacity check. It also contains a germination-test worksheet that averages four 100-seed replicates and checks the spread against the ISTA maximum tolerated range, so you can tell whether a test result is even reliable before you plan around it.

The stakes are ordinary arithmetic with expensive consequences. A grower who reads a 10 lb/acre pure live seed recommendation and orders 10 lb/acre of bulk seed at 63.75 percent PLS plants a stand roughly 36 percent thinner than the specification, which on a conservation seeding usually means weed pressure and a failed practice payment. A propagator who sows one seed per cell at 70 percent establishment ships trays that are 30 percent empty. Neither error is exotic; both are eliminated by doing the multiplication on this page before the seed is ordered.

One vocabulary note that causes constant confusion: "germination rate" is used in two different senses. In seed labeling and in this calculator it means a percentage — how many of 100 pure seeds produced normal seedlings. In seed physiology it often means a speed — how fast a population germinates, summarised by mean germination time or a germination index. Both meanings are explained below, but only the percentage sense feeds the seeding-rate arithmetic.

How to use this pure live seed and seeding-rate calculator

1. Copy the seed-lot block straight off the label or test report. Enter germination as the percentage of normal seedlings, enter hard or dormant seed on its own line if the label reports one, and enter purity as the pure-seed percentage by weight. If the label shows no purity figure — common on small vegetable packets — leave purity at 100 and read the result as a per-seed rather than per-pound answer.
2. Pick the crop if you want the entered lot compared with the federal minimum germination standard in 7 CFR 201.31. The comparison uses total germination, because the regulation states that the standards "shall be construed to include hard seed."
3. Set the stand target and the losses. Desired plants is the final stand you want at harvest or at shipping. The survival factor is the fraction of viable, germinable seed that becomes a usable plant in your conditions — the Forest Service defines it as seedlings produced divided by viable seeds planted, so 60 seedlings from a square foot sown with 100 seeds at 80 percent laboratory germination gives 60/80 = 0.75, i.e. 75 percent. The safety margin is a deliberate over-order on top of that, not a second survival estimate.
4. Set seeds per cell or hill if you are sowing plug trays, soil blocks or hills. Leave it at 1 for single-seeded precision drilling.
5. Optionally add a bulk seeding rate. Enter the recommended rate in pounds per acre of pure live seed and the area to be planted; the calculator converts it to bulk pounds using the label-basis PLS.
6. Optionally add plot dimensions and spacings to check whether the ground can physically hold the stand you are planning.
7. Optionally use the germination-test worksheet. Enter the number of normal seedlings in each of four 100-seed replicates; the worksheet reports the rounded average, the ISTA Table 5B maximum tolerated range, whether your replicates agree, and offers to push the average into the germination field.
8. Read the sensitivity table that appears under the result. It re-runs the whole plan across a band of germination percentages so you can see how much a five-point germination error actually costs in seed.
9. Save your work with the copy, download and shareable-link buttons. The link encodes every input, so a colleague opening it sees the identical scenario.

Formula set: from a seed label to seeds in the ground

Every output on this page comes from one of the six relationships below. Percentages are written as percentages; the code converts to decimals internally.

1. Germination percentage. The laboratory result is the mean of four replicates of 100 pure seeds, expressed as a percentage of normal seedlings and rounded to the nearest whole number:

with n_i the normal seedlings and N_i the seeds sown in replicate i. Under both AOSA and ISTA rules the four replicates total 400 seeds.

2. Total germination. Labels report hard seed (and, for many species, dormant or fresh seed) separately from germination. Total germination adds them back:

3. Pure live seed. The USDA NRCS pure live seed technical note gives the formula in exactly this form — purity times germination divided by 100:

The calculator reports this label-basis PLS and, alongside it, a prompt-establishment PLS that substitutes G for G_total because hard and dormant seed will not produce a plant in the current season.

4. Per-seed establishment probability. The Forest Service sowing-rate formula divides the desired stand by the product of germination, purity and a survival factor, which is the same as saying that each sown seed becomes an established plant with probability

The three factors multiply because they are sequential filters on the same seed: it must be the right species, it must produce a normal seedling, and that seedling must survive.

5. Cells and seeds to sow. A cell (or hill, or plug) sown with k seeds stays empty only if every one of the k seeds fails, so the probability of an empty cell is (1 − p)^k. Applying a safety margin B on top of the target stand D:

With k = 1 this collapses to the familiar D(1 + B/100) / p. The probability that a cell carries two or more seedlings and therefore needs thinning is

6. Bulk seeding rate. Recommendations from NRCS, seed suppliers and conservation specifications are quoted in pounds per acre of pure live seed, but seed is sold by bulk weight. The conversion is:

7. Plot capacity. The optional layout check treats each row as occupying a band equal to the row spacing and each plant as occupying a length equal to the in-row spacing, both converted from feet to inches:

8. Germination speed, for completeness. When a paper reports a "germination rate" that is a speed rather than a percentage, it usually means mean germination time — the seedling-weighted average of the counting days, where n_i seeds germinated on day t_i:

A lot can have a high germination percentage and a slow, drawn-out mean germination time; vigour tests exist precisely because those two properties diverge as seed ages. This calculator does not compute mean germination time — it needs daily counts, not a label — but the distinction matters when you compare published figures.

Worked example: a 200-plant tomato stand from a real seed label

A market grower is planning 200 harvestable field tomatoes. The seed lot report reads: germination 80 percent normal seedlings, hard or dormant seed 5 percent, purity 98 percent. Farm records from the last three seasons show that 90 percent of the seedlings that emerge in the propagation house survive hardening off, transplanting and early field pressure. The grower wants a 15 percent safety margin because the plants are committed to a market contract, and sows one seed per plug cell.

Step 1 — total germination and pure live seed. Total germination is 80 + 5 = 85 percent. Label-basis pure live seed is 98 × 85 / 100 = 83.3 percent. Prompt-establishment pure live seed, which drops the 5 percent hard seed because it will not move this spring, is 98 × 80 / 100 = 78.4 percent.

Step 2 — establishment probability. p = 0.784 × 0.90 = 0.7056, i.e. each sown seed has a 70.56 percent chance of becoming a field plant.

Step 3 — seeds to sow. With k = 1 the cell formula reduces to 200 × 1.15 / 0.7056 = 230 / 0.7056 = 325.96, rounded up to 326 seeds. Sowing 326 seeds should yield 326 × 0.784 = 256 normal seedlings and 326 × 0.7056 = 230 established plants, giving 30 plants of headroom over the 200-plant commitment.

Step 4 — what double sowing would change. At k = 2 the empty-cell probability falls from 29.44 percent to 0.2944² = 8.67 percent. The grower would need 230 / 0.9133 = 251.8, i.e. 252 cells and 504 seeds, and roughly 49.8 percent of those cells would carry two seedlings needing thinning. Whether that trade is worth it is an economics question: at $0.05 per seed the extra 178 seeds cost about $9, which is cheap insurance; at $0.50 per hybrid seed it costs $89 and single sowing plus a few spare trays wins.

Step 5 — the bulk conversion, on a field crop. Suppose the same grower is also drilling a cover crop with a specification of 10 lb/acre PLS across 4 acres, using a lot at 83.3 percent PLS. Bulk rate is 10 ÷ 83.3 × 100 = 12.00 lb/acre, so the order is 48.0 lb. The NRCS technical note works the identical arithmetic on a poorer lot — 85 percent purity and 75 percent germination gives 63.75 percent PLS, and a 10 lb/acre PLS recommendation becomes 15.69 lb/acre of bulk seed. Ordering 10 lb/acre of that lot would have delivered a stand 36 percent below specification.

Step 6 — plot capacity. The tomato block is 100 ft long by 40 ft wide, with rows 48 in apart and plants 24 in apart in the row. Rows = ⌊(40 × 12) / 48⌋ = ⌊10⌋ = 10; plants per row = ⌊(100 × 12) / 24⌋ = 50; capacity = 500 plants. The 200-plant target uses 40 percent of the block, so spacing is not the binding constraint.

Step 7 — was the germination test even reliable? The laboratory reported replicates of 88, 84, 91 and 86 normal seedlings per 100. The average is 87.25, which rounds to 87. ISTA Table 5B Part 1 tolerates a range of 13 percentage points at an average of 87; the observed range is 91 − 84 = 7. The replicates agree, so 87 percent is a reportable germination. Had the range been, say, 22, the rules require a retest rather than a report.

Reading the result: what each output number means

Pure live seed (label basis) is the number to use when you are converting a pure-live-seed seeding recommendation into bulk pounds, when you are comparing the price of two lots, or when you are checking a contract. Price per pound of bulk seed is close to meaningless; price per pound of PLS is the honest comparison, and a nominally cheap lot at 60 percent PLS is usually more expensive than a premium lot at 90 percent.

Pure live seed (prompt establishment) is the number to use when you are predicting this season's stand, because it excludes hard and dormant seed. On species with negligible hard seed the two figures are identical; on legumes and many natives they can differ by 10 to 30 points.

Establishment probability is the single per-seed number that everything else hangs on. Below about 0.5 the plan becomes fragile: small errors in any one factor swing the seed requirement sharply, which is exactly what the sensitivity table shows.

Seeds to sow already includes the safety margin. It is a planning quantity, not a physical guarantee; actual emergence is a binomial draw around the expectation, and for small stands the spread is proportionally larger.

Expected established plants is the mean outcome if you sow the recommended quantity. If it lands far above your target, your safety margin is doing more work than you intended and you are buying seed you will thin away.

Federal standard check compares the lot's total germination with the minimum germination standard for that crop in 7 CFR 201.31. A lot below the standard may still be sold in interstate commerce, but it must be conspicuously labelled "Below Standard" together with the germination and hard seed percentages.

Federal minimum germination standards for common vegetable seeds

Note how low several standards are. A carrot or pepper lot at 60 percent germination is perfectly legal, which is precisely why a seeding-rate calculation that assumes "seeds mostly grow" fails on those crops. The standards are floors for commerce, not targets for planning.

The ISTA replicate tolerance table used by the worksheet

The table is deliberately asymmetric: a lot averaging 95 percent must have replicates within 9 points of each other, while a lot averaging 60 percent is allowed a 19-point spread, because binomial sampling variance peaks near 50 percent. If the observed range exceeds the tolerance, ISTA requires a retest with the same or an alternative method; the first result is not reportable.

Storage, seed age and where germination goes when you are not looking

A germination figure is a snapshot with an expiry date. Seed deteriorates as a function of moisture content, temperature and time, and the classic seed-viability literature models that decline as a straight line on a probit scale, with both temperature and moisture entering the loss constant. The operational summary for growers is simple: dry seed stored cold lasts; damp seed stored warm does not.

Two cautions on the freezer row. First, seed must be dried before freezing; freezing seed at field moisture ruptures cells and destroys viability. Second, a few species are recalcitrant rather than orthodox and cannot be dried or frozen at all. Within vegetables, the short-lived families are the ones to watch: onion, leek, parsnip and parsley routinely lose most of their germination in a single year of poor storage, while tomato, cucurbits and brassicas are comparatively durable.

Species behaviour that will break a naive germination test

Temperature windows. Both AOSA and ISTA prescribe a specific substrate and temperature for each species precisely because germination is not temperature-neutral. Lettuce shows thermodormancy above roughly 75 F and will read as a dead lot if tested warm; peppers and melons barely move below 60 F. A test run at the wrong temperature does not measure viability, it measures the temperature.

Light requirements. Lettuce, celery and many small-seeded flowers are positively photoblastic and germinate poorly in darkness; a few species require darkness. Testing protocols specify the light regime, and burying a light-requiring seed too deep in the field produces the same false negative.

Hard seed and physical dormancy. Legumes and many natives carry impermeable seed coats. Those seeds are alive and count toward total germination and PLS, but they will not imbibe until scarified or weathered. This is the central reason the calculator separates label-basis from prompt-establishment PLS.

Physiological dormancy. Many perennials and tree species need cold-moist stratification, and the rules prescribe prechill treatments for them. A germination test run without the prescribed treatment reports dormancy as death.

Slow species. Parsley, carrot and parsnip may take 14 to 21 days under optimal conditions. Counting on day 7 and declaring a lot dead is a common and expensive error.

Limitations and assumptions you should hold in mind

• Laboratory germination is a ceiling, not a forecast. The test is run under the optimal substrate, temperature and moisture prescribed for the species. Field emergence is almost always lower, and the survival factor input is the only place where that reality enters the arithmetic. If you do not have your own history-plot data, treat any survival factor above 0.9 as optimistic.
• The three factors are assumed independent. Germination, purity and survival are multiplied as if they were independent filters. In practice a poorly cleaned lot often also germinates and establishes worse, so the product can be mildly optimistic for low-quality seed.
• Expected values are not guarantees. Emergence is binomial. For a 20-plant home planting at p = 0.7 the standard deviation is about 2 plants, so outcomes between 10 and 18 plants from 20 seeds are entirely ordinary. The safety margin exists to absorb that variance; it does not remove it.
• The cell model assumes independence between seeds in a cell. Real cells share moisture, temperature and disease, so empty-cell rates in practice run slightly above the binomial prediction.
• The plot capacity model is a simple grid. It assumes rectangular geometry, uniform spacing, and that each row and each plant occupies exactly its nominal band. It ignores headlands, wheel tracks, irrigation lines and bed edges, so treat the capacity figure as an upper bound and subtract your own access allowances.
• Purity is a weight fraction, germination is a count fraction. Multiplying them to get PLS is the standard convention used by NRCS and the seed trade, and it is exact when applied to weight-based seeding rates. If you are counting individual seeds out of a packet rather than weighing bulk seed, the purity term should normally be left at 100.
• Sample size limits confidence. A 400-seed official test has a standard error of roughly 2 percentage points near 85 percent germination. A 20-seed kitchen test has a standard error near 8 points, so it can distinguish "clearly good" from "clearly dead" and little else.
• Regulatory minimums are floors. The 7 CFR 201.31 standards permit lots that are marginal for planning purposes, and state seed laws may impose additional or different requirements. The comparison in this calculator is informational and is not a compliance determination.

Frequently asked questions about germination testing and pure live seed

Sources and standards behind this calculator