Data updated August 2026 from the compiled EPA AQS release in this database.

According to the U.S. Environmental Protection Agency (EPA), 977 counties reported Air Quality Index data through the federal Air Quality System (AQS) for the 2025 reporting year. The ranking below draws on those official records; see the methodology for how median AQI is computed.

Counties below the 100-day monitoring floor are excluded from this ranking: with only a handful of readings, a single bad-air stretch can collapse a county's Good-day share. Across all 951 floor-qualifying counties, the cohort median Good-day share is 77.5%; the first-ranked county records 8.8%. The gap therefore reflects a sustained annual pattern rather than a thin sample.

Filter this ranking

Filters keep the Good-day-share ranking contract. A state filter ranks that state's qualifying counties; pollutant filters keep rows whose annual day-type count is led by ozone, PM2.5, or PM10. Raising the monitored-days floor tightens the sample beyond the default 100-day Module-34 guard.

Showing 13 matching Arizona counties . Back to the national top 50.

Where the national worst cohort concentrates

Among the unfiltered national top 50 (lowest Good-day share, ≥100 monitored days), California contributes 13 counties , followed by Texas (11). Use the state filter above to rank every qualifying county inside one jurisdiction instead of truncating the national list.

Top contributing states in the worst-50

• California 13/50
• Texas 11/50
• Ohio 4/50
• Arizona 3/50
• New Mexico 2/50
• Arkansas 2/50
• Michigan 2/50
• Indiana 2/50

What leads the bad-air days

On each county row, EPA reports how many monitored days were ozone-, PM2.5-, or PM10-classified. In the national worst-50 cohort, PM2.5-led days are the plurality (32 of 50). Counts: ozone 16 · PM2.5 32 · PM10 2 . Filter by driver to keep only counties whose annual day-type count is led by that pollutant.

Ozone-led vs particle-led counties

Ozone-led rows (16) often sit in long warm-season photochemical corridors; PM2.5-led rows (32) more often track wildfire or winter particle episodes. Neither label is a health diagnosis - it is only which pollutant-day count dominated the annual monitor record.

How far below the cohort median

Across all 951 floor-qualifying counties, the median Good-day share is 77.5%. The #1 county here (Maricopa, Arizona) records 8.8% Good days - 68.8 percentage points below that median. That gap is why a thin sample is excluded: short monitor series can invent false extremes.

Annual record vs live advisory

This page answers “which counties had the worst annual Good-day share in 2025?” It does not answer “is the air safe today?” For current conditions and alerts use AirNow. State hubs at /states roll the same annual contract up one geography level.

Compare any two counties

Need a side-by-side annual record (Good-day share change + median AQI endpoints)? Use the compare form above or the county compare tool - it searches the full EPA AQS county set, not just this ranking slice.

Also useful nearby

• Best air quality counties - the inverse Good-day-share ranking
• AQI lookup - look up one county’s annual bands
• Methodology - how Good-day share and the 100-day floor are derived

Ranked counties

Reading the table columns

Good Days % is the ranking key. Median AQI breaks ties. Max AQI flags extreme days (values above 500 are beyond EPA’s standard scale). Day driver is the pollutant with the most classified days on that county’s annual row.