How do you estimate grain yield before harvest?

Count the heads in a measured area, count the grains in a representative sample of heads, and apply a thousand grain weight. Heads per square metre multiplied by grains per head gives grains per square metre, and multiplying that by the weight of one grain and by 10,000 gives grams per hectare, which converts to tonnes. The counting is the hard part: take several samples from different parts of the paddock rather than one from the gateway.

What is thousand grain weight?

The weight in grams of exactly one thousand grains, which is the standard way of expressing grain size because individual grains are too light to weigh usefully. It differs by crop, by variety and by season, and a dry finish will pull it down noticeably. Use your own figure from a sample if you have one, since a borrowed number is the largest single source of error in a pre harvest estimate.

How accurate is a pre harvest yield estimate?

Accurate enough to plan storage, cartage and cash flow, and not accurate enough to sell against. Every input is a sample, and the crop still has to finish, so a dry spell or a wet one moves the thousand grain weight after the count was taken. Treat it as a range rather than a number, and take more samples if the paddock is variable.

Grain Yield Estimator NZ

Count the heads, count the grains, weigh a sample, and this turns those three numbers into an estimate of what a cereal paddock will yield, in tonnes a hectare and across the whole block, with the value at your price. It is the arithmetic that lets storage, cartage and cash flow be planned before the header goes in rather than discovered afterwards. Be clear about what the estimate is worth. Every input is a sample, and the crop has not finished, so a dry spell after the count will pull the thousand grain weight down and a kind finish will lift it. That makes this a planning figure and not something to sell against. Two things improve it more than any adjustment to the maths. The first is sampling from several places across the paddock rather than one convenient spot near the gate, because the gateway is almost never representative and a variable paddock will not average itself. The second is using your own thousand grain weight from an actual sample, since a borrowed figure is the single largest source of error here: it differs by crop, by variety and by season, and it is the one input that carries the whole estimate on its back.

Advertise on this page

Heads a square metre

Grains a head

Thousand grain weight

Harvest loss allowance

Paddock area

Grain price

8.08 t/ha

tonnes a hectare, after the harvest loss allowance

Before losses8.51 t/ha

Whole paddock258.6 t

Grains a square metre18,900

Value$116,348.40

A thousand grain weight five grams lower would give 7.18 t/ha, which is 28.7 t off this paddock. A planning figure, not a selling figure. Every input is a sample and the crop has not finished.

No thousand grain weight, head count or loss allowance is supplied, because all three differ by crop, variety, season and header setup. Take the thousand grain weight from your own sample: it is the input that carries the estimate.

How it works

Heads a square metre multiplied by grains a head gives grains a square metre. The thousand grain weight divided by a thousand gives the weight of a single grain in grams, and multiplying through by 10,000 square metres in a hectare gives grams a hectare, which divides by a million for tonnes. The harvest loss allowance then takes off whatever you expect to leave behind the header. The whole paddock figure multiplies the yield by the area, and the value applies your price to that.

Sampling is the accuracy, not the arithmetic

The formula is exact and contributes no error at all. Everything that can go wrong is in the three counts. A head count taken beside the gateway will read high because of the extra nitrogen and the thinner competition there, a grain count from the best heads in the sample will read high for the obvious reason, and a thousand grain weight borrowed from a book rather than measured from the crop can be out by a tenth in either direction. Taking half a dozen samples from across the paddock and averaging them is worth more than any refinement to the calculation.

Worked example

A wheat crop with 450 heads a square metre, 42 grains a head and a thousand grain weight of 45 grams carries 18,900 grains a square metre. At 0.045 grams a grain that is 850.5 grams a square metre, or 8.505 tonnes a hectare before losses. A 5 percent harvest loss allowance brings it to 8.08 tonnes a hectare, and across 32 hectares that is 258.6 tonnes worth $116,348.40 at $450 a tonne. If the thousand grain weight turns out to be 40 rather than 45, the same counts give 7.18 tonnes a hectare, which is 28.7 tonnes and about $12,900 less off the same paddock.

Related calculators

Grain Moisture Shrink: what the load loses drying.
Acres per Hour: how long harvest will take.
Nitrogen Cap: the nitrogen behind the crop.
Farm Lease Rate: what the country cost to crop.
FX Drip-Feed vs Lump Sum Calculator NZ 2026/27: Free NZ calculator comparing converting a lump sum of NZD to USD in one transaction against spreading it across several.
Growing Degree Days Calculator NZ: Growing degree days from daily maximum, minimum and base temperatures, tracked to date and projected to a growth stag...

Data sources: the yield formula is arithmetic and holds for any cereal, so nothing here is a rate. Thousand grain weight, head and grain counts and the loss allowance all come from your own paddock and your own header.