CyberMax-tools/fipsbook-us-census-geographies

Fipsbook: every US Census geography with its FIPS code, centroid and land area (2025)

Need it fresh, filtered or via API? This free file is a snapshot (Census geographies and FIPS codes), last updated 2026-09-25.

• Coffermap FY2025 federal spending workbook ($15): every US county and state in one Excel file with top agencies, contractors and industries.
• Ziplore ZIP code API (free, no key): looks up any ZIP by API: radius search, nearest ZIP, city to ZIPs, demographics.
• Get an email when this dataset updates: free, double opt-in, unsubscribe any time.

Free lookup tables for joining, mapping and geocoding US data: 210,319 rows across 11 layers (states, 3,222 counties, 32,350 places, 36,427 county subdivisions, 33,791 ZIP code areas (ZCTAs), 85,396 Census tracts, 935 metro/micro areas, 440 congressional districts of the 119th Congress, 6,843 state legislative districts and 10,863 school districts), each with its GEOID/FIPS code, name, land and water area (m² and sq mi) and an internal-point latitude/longitude.

Built from the US Census Bureau 2025 Gazetteer Files and cleaned up so they are easy to use: typed columns (no trailing tabs, numbers as numbers), one Parquet table per layer, the state name on every row, and for tracts and county subdivisions the parent county FIPS and county name already joined in. Congressional districts get a readable label (NJ-07, WY-AL).

Have addresses, not FIPS codes? Pinloft — Batch Geocoder turns address lists into latitude/longitude plus the county, tract and block FIPS and the congressional, state legislative and school districts that join straight onto these tables ($0.0005 per matched address, misses free). Try it free first: Pinloft demo. Need ZIP code demographics, radius search or city → ZIP? Ziplore — US ZIP Code API (free lookup, free dataset).

Load it

from datasets import load_dataset
counties = load_dataset("CyberMax-tools/fipsbook-us-census-geographies", "counties", split="train")
tracts = load_dataset("CyberMax-tools/fipsbook-us-census-geographies", "tracts", split="train")

import pandas as pd
url = "https://huggingface.co/datasets/CyberMax-tools/fipsbook-us-census-geographies/resolve/main/data/{}.parquet"
cd = pd.read_parquet(url.format("congressional_districts"))
print(cd[cd.state == "NJ"][["label", "land_sqmi", "lat", "lon"]])

Layers (configs)

Config	Rows	GEOID format	Extra columns
states	52	2-digit state FIPS
counties (default)	3,222	5-digit state+county FIPS	ansi_code
places	32,350	7-digit state+place	lsad, funcstat
county_subdivisions	36,427	10-digit	county_fips, county_name
zctas	33,791	5-digit ZIP Code Tabulation Area
tracts	85,396	11-digit state+county+tract	county_fips, county_name, tract_code
metro_areas	935	5-digit CBSA code	cbsa_type (Metropolitan/Micropolitan), csa_fips
congressional_districts	440	4-digit state+district (119th Congress)	district, label
state_senate_districts	1,964	state+district
state_house_districts	4,879	state+district
school_districts	10,863	7-digit unified school district	lowest_grade, highest_grade

Common columns: geoid, geoid_full (Census fully-qualified id), name, state (USPS code), state_fips, state_name, land_m2, water_m2, land_sqmi, water_sqmi, lat, lon.

lat/lon are the Census internal point: a point guaranteed to be inside the area (usually near its centre), good for pins, labels and nearest-area lookups, not a population-weighted centre.

Good to know

• Connecticut uses its 9 planning regions as county-equivalents (FIPS 09110–09190) since 2022; the 8 old counties are not in the 2025 files. Pinloft maps both.
• ZCTAs are Census approximations of USPS ZIP code delivery areas; PO-box-only ZIPs have no ZCTA.
• Areas and codes are the Census Bureau's 2025 vintage; they change each January (districts after redistricting).

When NOT to use this

• You need boundaries (polygons): use Census TIGER/Line shapefiles. This dataset has one point per area.
• You need population or income: join these codes to American Community Survey tables (for ZIPs, see Ziplore's dataset above).

Source and licence

US Census Bureau, 2025 Gazetteer Files (https://www.census.gov/geographies/reference-files/time-series/geo/gazetteer-files.html). Works of the US federal government are in the public domain; this cleaned compilation is released under CC0 1.0. Not endorsed by the Census Bureau. Refreshed yearly when a new Gazetteer vintage is released.

Published by CyberMax.

More free from CyberMax

• All free CyberMax demos & datasets
• Pinloft: Batch Geocoder with Census FIPS & Districts
• Ziplore US ZIP Codes with Census Demographics (41,682 ZIPs)

Quickstart notebook

Load this dataset with pandas and see real outputs in the quickstart notebook.