12 Free Sources for US Parcel Data (and What Free Actually Costs)
Most US counties publish parcel data for free. Of the 3,164 counties we track, we still could not source usable parcel data for roughly one in six — and the ones that do publish each use a different schema. Here is where to look, and what the hunt costs you.
Most counties in the US publish parcel data, and a lot of it is free. The
difficulty was never availability. It is that every county publishes on its own
schedule, in its own format, with its own field names, and nothing tells you
which of the three thousand you need are worth the trip.
We know the shape of that problem in some detail, because we did it county by
county. Of the 3,164 counties we track coverage for, about one in six had no
usable parcel data we could source at all — no download, no API, no usable
export, or a portal that returns a viewer and nothing else. The other five in six
each needed their own mapping from local field names into a common schema.
This guide is the shortcut to the sources worth your time, the honest version of
what free costs, and where paying stops being a waste of money.
What free parcel data actually gets you
Free county data is authoritative — it comes from the office that maintains the
record. What it does not come with is consistency. PARCELNBR in one county isPIN in the next and TAXPIN in the one after that. Acreage arrives as a string
with a comma in it. Assessed value is sometimes in dollars and sometimes in
hundreds. Geometry arrives in whatever projection the county's GIS vendor
defaulted to in 2013.
For one county, that is an afternoon. For thirty, it is a project. For all of
them, it is what we do full time.
Tier 1 — Work out who has what
1. USGS Authoritative Source Parcel Viewer.
The USGS maintains an index of which jurisdictions publish parcel data and where.
It is a secondary source, so do not download from it — use it to find the
authoritative portal, then go straight there.
Best for: establishing whether your county publishes at all.
2. Data.gov. Filter for "parcel" or "cadastral" plus
your state. Quality ranges from current to badly stale, and the listing rarely
tells you which, so check the metadata date before you commit.
Best for: obscure local datasets that do not surface in a web search.
3. Open Data Network. Many counties and
cities run Socrata portals, which means a real API and a predictable CSV or
GeoJSON export instead of a zip behind a form.
Best for: counties with a modern open-data programme — usually far quicker than
a legacy assessor site.
Tier 2 — Go straight to the county
4. County assessor and GIS portals. The authoritative source, and the
freshest. These search patterns find them faster than the county's own
navigation:
[county name] assessor parcel shapefile download[county name] GIS open data[county name] arcgis rest services parcels
That last one is the useful one. Most counties running ArcGIS Server expose a
REST endpoint that will hand you GeoJSON directly, even when the website only
offers a viewer. If you find /arcgis/rest/services/, you can usually query it.
Best for: maximum freshness and the complete attribute table.
Tier 3 — Statewide programmes that actually work
A handful of states solved this properly and publish one normalised parcel layer
covering every county. If your project is in one of these, start here and skip the
county-by-county hunt entirely.
5. Wisconsin Statewide Parcel Map.
The best statewide parcel programme in the country, and not a close contest.
Every county, one schema, annual releases, documented field definitions, free.
Best for: any Wisconsin work — and as the example to point at when you ask
another state why theirs is worse.
6. Montana Cadastral. Statewide parcels
with ownership, maintained by the Montana State Library. Downloads and a usable
API.
Best for: large rural areas where county portals are thin.
7. NC OneMap. North Carolina
aggregates all 100 counties into a statewide parcel layer with a documented
schema.
Best for: multi-county North Carolina analysis.
8. Utah UGRC. Utah's
geospatial resource centre publishes statewide parcels through the SGID, with
clear documentation of what each county contributed.
Best for: Utah, and a good model of provenance done properly.
9. New York State tax parcels. NYS publishes
statewide tax parcel data, county by county, on a documented annual cycle.
Best for: New York outside the city — NYC has its own, and MapPLUTO is better
for it.
10. Minnesota Geospatial Commons. Statewide parcel
data with genuinely good metadata, including which counties are current.
Best for: Minnesota, and for seeing what honest coverage documentation looks
like.
11. Oregon ORMAP. A statewide cadastral programme
funded by a recording fee, which is why it exists at all when some neighbouring
states' do not.
Best for: Oregon parcels and taxlot boundaries.
Also worth knowing, in the same category: the
Alaska Open Data Geoportal,
AZGEO (Arizona, free registration), the
California State Geoportal — which replaced the
retired Cal-Atlas — the Florida Geographic Data Library,
and Texas's TxGIO. Two university guides
are quietly among the best indexes anywhere:
Washington University's
and Purdue's, both
organised by state.
Tier 4 — Federal land and the survey grid
12. BLM cadastral data.
Township, range and section grids plus federal land boundaries for the western
states. Not parcels in the assessor sense, but if your work touches federal land,
grazing, or mineral rights, you need the PLSS underneath.
Best for: federal land, and anything that needs the survey framework rather than
tax parcels.
What free actually costs
Not money. Time, in five places.
Schema normalisation. Budget four to eight hours per county the first time,
less once you have a pattern for that vendor's exports. Thirty counties is a
month of work that produces no product.
Update tracking. Nothing notifies you. You need a job that checks each source
and a way to tell a real update from a republished file with a new timestamp. Some
counties refresh weekly, some annually, some when someone remembers.
Coverage gaps. About one in six counties has nothing usable, and they cluster
in rural areas — which is exactly where you need coverage for infrastructure,
energy, land and agricultural work.
Quality. Misaligned geometry, self-intersecting polygons, addresses in the
owner-name field, duplicate parcels from a bad join. Free sources have no budget
for QA, and the errors are not evenly distributed: one county will be immaculate
and the next will have a few per cent of its parcels stacked on a single
centroid.
Format conversion. Proprietary geodatabases, CAD files, and the occasional
county that publishes a PDF map book and considers the job done.
When to pay
Free is the right answer for a single county, a visual reference, an academic
project with time to spend, or a prototype before there is a budget.
Pay when you need consistency across many counties, when something in production
depends on the data, when your team's hours are worth more than the licence, or
when you need to answer "how fresh is this, exactly?" for a client.
What we do differently
Two things, and only one of them is the obvious one.
One schema across 3,164 counties. parcel_id, owner_name,assessed_value, acreage, land_use, zoning, geometry — the same field names
everywhere, with each county's original codes kept alongside our normalised
categories so you can audit any mapping we made. Every delivery includes five
formats — GeoParquet, GeoPackage, Shapefile, GeoJSON and CSV — and a README
describing every field.
We publish where the data is thin. Every county page on our site shows that
county's actual field fill rates and the vintage of its source. If a county's
assessor does not publish year built, our page for it says so before you buy
rather than after. It is the only way for you to know whether the fields your
model needs exist where you need them — and it is why the "one in six" number
above is on this page at all.
Pricing follows from that. $49 buys any area you draw, up to 5,000 parcels. $199
buys an entire county, whatever its size. The API starts at $29 a month. A free
sample returns real GeoJSON for a small area with no account and no card.
Test before you scale
Download one county. Load it in your own stack. Check the geometry registers
against your basemap, check the fields you need are populated rather than
present-and-empty, check the update date is what the portal claimed.
Parcel data is foundational, and bad foundations fail under production load
rather than during evaluation. Whether you go free or paid, spend an afternoon on
one county before you build anything on three thousand.
Check per-county coverage and field fill rates at
getparceldata.com/coverage, or draw an area
and buy it outright at getparceldata.com/extract.