What USDA forecasts it will cost to grow nine US field crops in 2026

The USDA Economic Research Service publishes what it expects it will cost to grow a crop, per planted acre, for the United States as a whole. AgricultureID captured the current release on 27 August 2026. Here is what it says for 2026.

Crop	Total costs listed	Operating costs
Rice	$1,411.44	$838.30
Groundnut	$1,223.94	$666.95
Cotton	$974.12	$583.82
Maize	$935.79	$473.03
Soybean	$683.89	$257.43
Barley	$526.92	$206.19
Oats	$523.03	$186.85
Sorghum	$466.20	$190.93
Wheat	$419.41	$169.08

All figures in US dollars per planted acre, as published.

Three things these numbers are not

They are not what anyone spent. ERS marks the years 2026F and 2027F. The F is the whole of what separates this dataset from an accounting of actual farm expenditure. A season that has not finished cannot be accounted for.

They are not about anywhere in particular. The region on every row is "U.S. total". ERS publishes regional breakdowns separately. A national average across a crop grown from the Mississippi Delta to the northern Plains is a real statistic about a real population of farms and is not a statistic about any of them.

They are not a margin. This is the one worth being blunt about. The dataset contains costs and no revenue. Divide total cost by an assumed yield and multiply by an assumed price and you have a break-even figure — but the yield and the price are yours, not USDA's, and the resulting number is a scenario wearing a statistic's clothes.

AgricultureID holds 24,916 farm economics figures and publishes no gross margin, no net return and no break-even price, for exactly this reason. Those three are recorded in the platform as absent, each with the reason it is absent.

Reading the split

The gap between total costs and operating costs is the allocated overhead: capital recovery of machinery and equipment, the opportunity cost of land, the opportunity cost of unpaid labour, taxes, insurance and general farm overhead.

For wheat, operating costs are 40% of the total. For rice they are 59%. That difference is not a statement about efficiency — it reflects how much of each crop's cost structure ERS allocates to land and machinery rather than to inputs bought within the season.

Note also that the totals here are ERS's own. AgricultureID does not add the cost items up to produce one, because the source decides what belongs in a total and an independently computed total would disagree with it for definitional reasons that would read as a data error.