25 Oct 2026 5 min read

Solar and farmland: what the USDA numbers actually say

The USDA solar and farmland figures are real, but their dates, geography and definitions matter. A sourced guide to acreage, development and project review.

By Dalton Barron, 337.news
Fact sheet • Sources checked October 25, 2026

The often-cited 336,000-acre rural solar estimate describes 2020. The 880-million-acre farm inventory comes from the 2022 Census of Agriculture. Comparing those figures gives a sense of national scale, but it does not measure the share of farmland converted to solar or establish the effects of a particular project.

Readers checking solar and farmland claims need the year, geography and definition behind each number. They also need to separate observed changes from modeled future scenarios.

What the 336,000-acre estimate measures

USDA’s May 2024 Economic Research Service report estimated a 2020 rural solar footprint of 336,090 acres in the contiguous United States. Its calculation used 7.5 acres per megawatt. The footprint includes panels, spacing between rows and a surrounding buffer. It excludes additional structures such as transmission lines.

This is an estimated rural solar footprint across land types. It is not a count of agricultural acres converted, a panel-surface measurement or a current national total. The report’s definition also differs from a developer’s entire leased or permitted project boundary.

What dividing by 880 million does, and does not, show

USDA’s 2022 Census of Agriculture, Table 1, reports 880,100,848 acres of land in farms in 2022 and 900,217,576 acres in 2017. Subtracting those figures gives a decline of 20,116,728 acres, about 20.1 million acres.

Dividing 336,090 by 880,100,848 and multiplying by 100 gives approximately 0.038 percent, which rounds to 0.04 percent. That calculation mixes a 2020 rural solar estimate with a 2022 U.S. farm inventory.

The numerator includes solar on nonagricultural land, while the denominator includes land in farm operations nationwide. The ratio therefore cannot be labeled the percentage of American farmland converted to solar. A conversion estimate would need matching geography, dates and a measurement of agricultural land actually affected.

A farm inventory decline is not a count of paved acres

USDA’s definition of land in farms includes crop, pasture and grazing land, along with qualifying woodland, other land within farm operations and land in conservation programs. It is broader than harvested cropland.

The Census totals show a net change in that inventory. They do not identify the destination or cause of every acre leaving it. The 20.1-million-acre decline cannot, by itself, be attributed to solar, housing or paving.

A University of Illinois farmdoc daily analysis published September 18, 2024, found a 74.7-million-acre decline in land in farms between 1997 and 2022, with pastureland accounting for 88 percent of that decrease. That is a different period from 2017–2022. It illustrates why farmland loss and cropland loss should not be treated as interchangeable measures.

Solar grew after the 2020 snapshot

The Energy Information Administration reported that developers added 31 gigawatts of utility-scale solar photovoltaic capacity in 2024, increasing total U.S. utility-scale solar capacity by 34 percent. Those are actual additions reported in EIA’s November 10, 2025 analysis, rather than the larger amount developers initially planned.

Generating capacity and land area are different measures. This later growth is a reason to date the 2020 acreage figure clearly. It is not enough to calculate today’s footprint by multiplying every new megawatt by a single historical acreage factor.

A small national footprint can overlap with agricultural land

USDA’s May 2024 regional summary says roughly 70 percent of rural solar projects installed between 2009 and 2020 were on agricultural land. That is a share of projects, not a measured share of national farmland converted.

The underlying report classified 43 percent of solar sites as cropland and 21 percent as pasture-rangeland before installation. These categories do not include every agricultural land-cover combination. They should not be added together and presented as a replacement for the broader agricultural total.

Those findings describe historical siting patterns. They do not identify the soils, drainage effects or agricultural production displaced by a specific development.

Future solar acreage figures are scenarios

The Department of Energy’s solar siting page, summarizing the Solar Futures Study, describes about 5.7 million acres of ground-mounted solar by 2035 and as much as 10 million acres by 2050. DOE compares those figures with approximately 0.3 percent and 0.5 percent of the contiguous United States.

Those percentages use all contiguous U.S. land as the denominator, not farmland. They are modeled future land requirements under the study’s assumptions, not measurements of current acreage or guarantees of future development.

Other development pressures use different measurements

American Farmland Trust’s historical analysis estimated that development lost or compromised about 2,000 acres of farmland and ranchland per day during 2001–2016. This is a historical average, not a newly measured daily rate for 2026.

The Trust’s business-as-usual scenario projects 18.4 million acres converted from 2016 to 2040, based on past development patterns. That figure covers the full scenario period. It should not be described as 18.4 million additional acres starting today.

These findings document development pressure, but they are not directly comparable with a one-year stock estimate of rural solar acreage. A claim ranking causes of farmland loss needs a common time period, geography and definition of conversion.

Agrivoltaics exists, but needs a project-specific plan

DOE explains that most large ground-mounted solar systems occupy land used only for solar energy production, while co-location with crops, livestock or pollinator habitat is possible. Outcomes depend on the design and agricultural activity.

A November 2025 Kleinman Center analysis reported that an early-2025 NREL database snapshot identified 596 agrivoltaic sites covering 65,699 acres, with 10,473 megawatts of capacity. The analysis notes that projects were self-reported and likely undercounted. Capacity in megawatts is not annual electricity production.

That inventory establishes examples of dual use. It is not a complete census, a current 2026 total or proof that a proposed site will produce crops. Pollinator habitat also should not be described as crop production.

What to ask about a specific lease or project

NYSERDA’s landowner guidance emphasizes evaluating lease terms and site characteristics. USDA NRCS conservation guidance addresses resource protection during solar siting, construction, operation and decommissioning. These are useful starting points for reviewing the actual documents.

A landowner’s consent to a lease is one part of that review. The agreement’s duration, extensions, agricultural permissions and restoration obligations need to be read together. Effects on neighbors require their own evidence.

How much land is inside the lease and project boundary, and how much is used for arrays, roads, substations and transmission?
What was the previous land use, what soils are present, and what agricultural production would change?
What does the drainage plan require, and who must address runoff problems?
What are the initial lease term, renewal rights and provisions governing sale, inheritance or transfer to another operator?
What removal and soil-restoration work is required, who funds it, and what happens if the operator cannot perform?
If grazing, crops or habitat are promised, what written plan governs planting, access, panel clearance, maintenance and monitoring?
Which permits and final ordinances apply, and what complaint, enforcement and appeal procedures do they provide?

These questions are a review checklist, not a statement that every jurisdiction already requires each item. National statistics provide context. The records for a particular site establish what is proposed and what is required.

Sources

Accessed October 25, 2026. Dates and limitations appear beside the relevant claims.

Updates and corrections: Sources checked October 25, 2026. Material corrections will be recorded here.