Methodology & Sources
CountyAtlas is built from official primary sources (the U.S. Census Bureau, the Bureau of Labor Statistics, and Wikidata) under one rule: every number is verified or omitted, never fabricated. This page documents where each figure comes from, the vintage it reflects, and the decisions behind it. It is the reference for the maps, tables, and downloadable dataset.
How many counties are in the United States?
There are two defensible answers, 3,143 and 3,144, and the difference is Connecticut. Its 8 counties stopped functioning as governments in 1960 but remained the Census Bureau’s statistical units until 2022, when the Bureau adopted Connecticut’s 9 councils of governments, its planning regions, as county-equivalents, with public data and boundary products reflecting the change from late 2022 (Federal Register notice 2022-12063, June 6, 2022). Counting the 9 planning regions yields 3,144 counties and county-equivalents; counting the 8 obsolete counties yields 3,143.
CountyAtlas uses the current basis of 9 planning regions, so its national total is 3,144. Sources that still report 3,143 are using Connecticut’s legacy-county geography; both numbers are correct for their vintage, and neither is a typo. The count also includes county-equivalents that are not called counties: Louisiana’s parishes, Alaska’s boroughs and census areas, and the 41 independent cities (38 of them in Virginia) that belong to no county. U.S. territories (Puerto Rico, Guam, and others) have their own county-equivalents and are counted separately, not in the 3,144.
Data sources & vintages
Every figure carries the vintage of its source. This table is generated from the site’s source registry. Every “Source:” line on the site renders from the same record, so a refresh changes it everywhere at once. The county-boundary vintage is read from the TIGERweb layer itself at build time: January 1, 2026.
Unemployment: county rates come from the BLS county workbook (“labor force data by county, latest 14 months”), retrieved Nov 2026. The newest month, September 2026, is marked preliminary by BLS and may be revised, so every figure from it carries “preliminary” beside its month.
All of these are freely usable public sources. The published dataset contains federal and Wikidata data, plus the county seats Wikidata lacks from the pages cited in the seat file, with no proprietary or license-restricted inputs, so it is freely redistributable.
Suppression: why some values are blank, not guessed
CountyAtlas never prints a survey estimate it can’t stand behind. For each one we compute the coefficient of variation (CV) from its margin of error (standard error = MOE ÷ 1.645, then CV = standard error ÷ estimate) and suppress any estimate whose CV reaches 0.30, the threshold in the Census Bureau’s Statistical Quality Standard F1-6.
The gate rarely fires. On the current data, 14 of the 3,144 counties suppress at least one field, and that is by design. For the headline figures we use sources built for small-area reliability: SAIPE for income and poverty, and PEP for population (see below). Those clear the bar almost everywhere. The suppressions that remain are almost entirely educational attainment (the one field drawn from raw 5-year ACS) in the smallest counties, where even a five-year sample runs thin.
The rule has no exceptions. Where an estimate is too noisy to trust, the cell is shown as suppressed, never as a blank that reads like zero and never as an invented value. Sentinel codes returned by the source APIs (negative placeholders that mean “no data”) are treated as missing, not as real numbers.
Why PEP, not ACS, for population
Two Census programs report county population, and they answer different questions. The Population Estimates Program (PEP) gives a single, timely estimate for a specific date (Vintage 2025 (July 1, 2025)). The American Community Survey gives a 5-year average (5-year 2020–2024) with a margin of error. For a plain “how many people live here” count, PEP is the more current and more direct answer, so it is the population shown throughout the atlas. We still use ACS where only it has the field (median age and educational attainment), and there the CV gate above applies.
ZIP to county: the postal answer, not the geometric one
A ZIP code is a U.S. Postal Service construct: a set of mail-delivery routes, not a shape on a map. So “what county is this ZIP in?” is a question about where the mail goes. CountyAtlas answers it with the HUD-USPS ZIP-County crosswalk, built from USPS delivery-point data, and reports each county’s share of the ZIP’s residential addresses.
This is deliberately not the Census ZIP Code Tabulation Area (ZCTA), a polygon the Census draws to approximate a ZIP. The two can disagree. ZIP 10505 (Baldwin Place, New York) has 60% of its residential addresses in Putnam County by the postal crosswalk, but its 2020 ZCTA lies entirely inside Westchester County. HUD-USPS 2026 Q2 · Census TIGER ZCTA 2020 Tools built on the ZCTA report the wrong primary county; the postal crosswalk reports where the addresses actually are.
City names and the counties a city reaches
Cities, towns and census-designated places come from the Census Bureau’s TIGERweb places, which record each place by its legal name. Where that name isn’t the one people use, the site shows the everyday name by rule. A consolidated city-county shows its city: Nashville for “Nashville-Davidson metropolitan government (balance)”, Lexington for “Lexington-Fayette”. A legal name with the common one in brackets shows the common one: Ventura for “San Buenaventura (Ventura)”. Honolulu stands for the Census Bureau’s “Urban Honolulu”, and New York City for the city named New York. Boise drops the “City” of “Boise City” because the Postal Service names every ZIP area that lies mostly inside the city “Boise” HUD-USPS 2026 Q2 · Census 2020 ZCTA–place file. The tools find a place by either name, with Saint or St., Fort or Ft. and Mount or Mt.
A city can reach into more than one county. The counties come from the American Community Survey, which counts every place by county part Census ACS 2020–2024 place parts. A city is listed first in the county its center falls in (the Census internal point), then in the others. New York City reaches five counties, one for each borough NYS Department of Health.
Balance populations. When a city and its county merge into one government but other towns inside keep their own, the Census Bureau “creates a separate place (balance) record for the portion of the consolidated city not within any other place” Census geography glossary. A (balance) population, such as Nashville’s or Indianapolis’s, therefore leaves out the separately incorporated places inside the consolidated city.
The principle: verified or omitted
One rule governs everything above: a rendered sentence exists only when a real field backs it. National totals are computed from the data, not typed in; a figure with no reliable source produces no claim rather than a plausible-looking guess; and the boundary basis (which Connecticut geography, which vintage) is stated, not assumed. That discipline is what makes the atlas and its dataset citable.
The county seats follow the same rule as the suppressed estimates. Wikidata records a seat that validates against a Census place inside the county for 3,052 of the 3,144 counties (97%, queried 2026-11-11). For 14 more, the seat comes from a page published by the county or the state, except for 3 where no government page could be reached and the row cites the Virginia Association of Counties. The dataset’s county_seat_source column carries each page’s address, and the seat file quotes the line that names the seat. The remaining 78 show the reason instead of a seat, never a blank. Of those, 66 have none: 41 independent cities, 11 Alaska census areas, 1 Alaska borough, 9 Connecticut planning regions, the District of Columbia and 3 counties run by the state or by a neighboring county. The 7 consolidated governments, where a city and its county or borough are one government, have no separate seat. New York City’s 5 boroughs have no official county seat. Nothing is inferred from the largest city.
Calculated values and their formulas
Some figures on the site are not in any source file: they are calculated from sourced values. Each is labeled “calculated” where it appears and links here. Inputs are always the sourced values above, after the CV gate. A suppressed estimate is excluded from every calculation, never imputed.
- Population density
- population ÷ land area (people per sq mi). Inputs: Census PEP V2025 · Census TIGER 2026.
- State totals
- sum over the state’s counties (population, land area, water area). Inputs: Census PEP V2025 · Census TIGER 2026.
- Averages per county
- state total ÷ number of counties. Inputs: Census PEP V2025 · Census TIGER 2026.
- Medians across counties
- middle value of the county values (mean of the two middle values for an even count). Inputs: Census PEP V2025 · Census SAIPE 2024.
- Spreads
- highest county value ÷ lowest county value. Inputs: Census PEP V2025 · Census SAIPE 2024.
- Income tiers vs poverty
- mean poverty rate of the N highest- and N lowest-income counties, N = min(10, ⌊counties ÷ 2⌋), only for states with ≥ 6 counties. Inputs: Census SAIPE 2024.
- Rankings and superlatives
- highest / lowest county value; suppressed values excluded. Inputs: Census PEP V2025 · Census SAIPE 2024 · Census ACS 2020–2024 · BLS LAUS September 2026, preliminary · Census TIGER 2026.
- Neighbors and border length
- counties sharing a boundary arc are neighbors; border length = shared arc length after clipping coastal and Great Lakes water. Inputs: Census TIGER 2026 · Census TIGER hydrography 2026.
- Square miles
- TIGER AREALAND or AREAWATER (m²) ÷ 2,589,988.11, rounded to 2 dp. Inputs: Census TIGER 2026.
Land area: why our figure can differ from QuickFacts
County land and water area come from the current TIGERweb county layer (January 1, 2026 vintage), converted from square meters to square miles. Census QuickFacts publishes “land area in square miles, 2020”, which comes from the 2020 Census Gazetteer file. For most counties the two agree. Where the Census Bureau has since reclassified land and water (common in Alaska, where hydrography has been redrawn), they don’t.
Yukon-Koyukuk Census Area, Alaska is the largest example by area. TIGERweb gives 145,767.54 sq mi of land and 2,110.61 sq mi of water; the 2020 Gazetteer gives 145,575.55 sq mi of land and 2,295.29 sq mi of water. Total area barely moved (147,870.84 → 147,878.15 sq mi); about 192 sq mi was reclassified from water to land, a shift that appears between the 2023 and 2024 Gazetteer files. We use the current TIGERweb value because it matches the boundaries we draw and the geography the population estimates use. Census TIGER 2026 · Census Gazetteer 2020
The dataset
Every county’s core fields are published as a single documented CSV of 3,144 rows, with federal and Wikidata data and the cited county seats. See the data page for the column dictionary, download, and version history.