USDA NASS Crop Yield Data API

USDA NASS API limitations

The USDA NASS Quick Stats API is the official source for crop statistics — and it's genuinely useful. But working with it directly has real friction for developers and analysts building production applications:

• Complex query structure. Every request requires specifying multiple parameters — commodity_desc, statisticcat_desc, unit_desc, prodn_practice_desc, util_practice_desc — and the valid combinations aren't always obvious. Getting corn grain yield (vs. silage, vs. sweet corn) requires knowing exactly which parameter combination NASS uses.
• Inconsistent county identifiers. NASS county data uses a combination of state ANSI code and county ANSI code rather than the 5-digit FIPS codes that USDA FSA, RMA, FEMA, and most other government systems use. Joining NASS data to any other county dataset requires a crosswalk table.
• No trend analytics. NASS returns the raw yield number. Whether that yield is above or below the county's long-run trend — the most analytically useful question — requires fitting your own OLS regression per county per crop.
• Rate limits and reliability. The NASS Quick Stats API has documented rate limits (1,000 requests/day on the public tier) and periodic downtime. Building a production system that depends on direct NASS API calls requires caching and error handling that most teams don't want to maintain.

R2Data2 solves all of these. One normalized endpoint, FIPS-keyed, with trend deviation pre-computed.

What R2Data2 adds over raw NASS

Supported crops

Yields are in the native NASS units: bushels/acre for corn, soybeans, wheat, and sorghum; pounds/acre for cotton. The yield_unit field in every record makes the unit explicit.

The yield_vs_trend_pct field

For each county and crop, R2Data2 fits an OLS trend line through the full NASS yield history. The yield_vs_trend_pct field then expresses each year's yield as a percentage deviation from that trend:

# yield_vs_trend_pct = (actual_yield - trend_yield) / trend_yield * 100
# Positive = above trend (good year); negative = below trend (stress year)

This is the metric that actually answers the question analysts care about: not "was corn yield 172 bu/ac?" but "was that 172 bu/ac above or below what this county typically produces at this point in its yield trajectory?" A county at –18% yield vs. trend in a D2 drought year is a very different story from one at –3%.

Use cases

Crop insurance: validating drought-year yield loss

MPCI and APH-based policies pay indemnities when actual yields fall below the guarantee. The yield_vs_trend_pct field provides a normalized, county-level yield deviation that cross-validates adjuster estimates with the historical NASS record. A county showing –20% vs. trend in a D3 drought year is consistent with material loss; a county at –2% in the same drought year warrants a closer look at why the deviation is so small.

Agricultural lending: collateral value benchmarking

Farm loan underwriters use county yield history to benchmark expected production value on collateral. A 10-year yield history with yield_vs_trend_pct immediately shows whether a county's production is consistently above trend (strong collateral), highly variable (higher risk), or trending downward (structural concern). Pull alongside RMA loss history for a complete picture.

Precision ag and crop analytics platforms

County yield data is a key validation dataset for crop models and field-level yield prediction. R2Data2 makes it trivial to pull a 20-year corn yield time series for any county, join it to NDVI anomaly or soil moisture data from the same API, and quantify the historical relationship between satellite-observed stress and realized yield outcomes.

Market research and commodity analytics

The national snapshot endpoint returns yields for every county that reported a given crop in a given year. For corn in a drought year, that's 1,200+ counties in one API call — useful for building county-weighted national yield estimates, mapping yield deviation geographically, or feeding commodity price models with county-level production data.

API endpoints

All crop endpoints require an API key in the X-API-Key header. The snapshot endpoint requires a Growth or Pro plan.

County yield history

Returns annual yield and harvested acreage for a county, with yield-vs-trend deviation. Optionally filter by crop; returns all supported crops if omitted.

# Corn yield history for Adair County, Iowa (FIPS 19001)
curl "https://r2data2.com/crops/county/19001?crop=CORN" \
  -H "X-API-Key: agr_your_key_here"

{
  "fips": "19001",
  "crop": "CORN",
  "records": [
    {
      "fips": "19001",
      "year": 2023,
      "crop": "CORN",
      "yield_value": 152.0,
      "yield_unit": "BU / ACRE",
      "acres_harvested": 114000,
      "yield_vs_trend_pct": -14.2
    },
    {
      "fips": "19001",
      "year": 2022,
      "crop": "CORN",
      "yield_value": 187.5,
      "yield_unit": "BU / ACRE",
      "acres_harvested": 118000,
      "yield_vs_trend_pct": 4.8
    }
  ]
}

Use start_year and end_year to scope the range. Omit crop to return all five supported crops in a single call — useful for building a multi-crop county profile.

Latest year — all crops

Returns the most recent NASS data year for every crop available in a county. A fast way to get current-season yields across all crops without specifying a year.

# Most recent year, all crops, for Adair County, Iowa
curl https://r2data2.com/crops/county/19001/latest \
  -H "X-API-Key: agr_your_key_here"

National snapshot — all counties for a crop and year

Returns yields for every county that reported a given crop in a given year. Corn in 2023 returns ~1,200+ county records — everything you need for a national yield map or production-weighted analysis.

# All county corn yields for 2012 (historic drought year)
curl "https://r2data2.com/crops/snapshot/2012?crop=CORN" \
  -H "X-API-Key: agr_your_key_here"

{
  "year": 2012,
  "crop": "CORN",
  "count": 1247,
  "records": [
    {
      "fips": "19001",
      "year": 2012,
      "crop": "CORN",
      "yield_value": 97.0,
      "yield_unit": "BU / ACRE",
      "acres_harvested": 108000,
      "yield_vs_trend_pct": -43.6
    }
  ]
}

See the Data Dictionary for the full field reference and API Docs for interactive testing.

Frequently asked questions