In this Article
Property data looks abundant: listings are visible on dozens of portals, and prices are discussed everywhere. Building anything on it runs immediately into a structure most other data categories do not have, where the authoritative source is a membership organisation rather than a vendor or a government body.
This guide explains how MLS access actually works, what the RESO standards do and do not solve, which sources are genuinely open, and where the contractual lines sit.
Key Facts
- There is no single MLS. Hundreds of regional multiple listing services operate independently, each with its own membership rules and data agreements.
- MLS access is membership-gated, not purchasable. Feeds are granted to licensed participants under agreements that constrain display and redistribution.
- RESO standards make the fields comparable, not the access. Two MLSs using the same standard still require two separate agreements.
- Public records are the open alternative: county assessor and recorder data is public in most US jurisdictions and is where valuation models start.
- Portal terms almost universally prohibit scraping listings, which is a contractual matter regardless of the page being publicly viewable.
Where does property data come from?
Four sources with very different access models. We call it the 4-part property data model.
| Source | What it holds | Access |
|---|---|---|
| 1. MLS feeds | Active listings, status changes, sold history, agent data | Membership and a data agreement; not sold to the public |
| 2. Public records | Ownership, parcels, assessed value, transfers, permits | Public in most US counties; formats vary enormously |
| 3. Portals and brokerages | Listings as displayed, plus their own estimates | Public pages, but terms almost always prohibit extraction |
| 4. Commercial aggregators | Combined and normalised datasets | Licensed; coverage and freshness vary by county |
The split that matters: rows one and three are contractually restricted, and row two is genuinely open. Most defensible products are built on public records plus a licensed feed, not on portal pages.
How does MLS access actually work?
Through membership, then through a per-MLS data agreement.
Each regional MLS decides who may receive a feed, usually licensed brokers and agents and the technology vendors they authorise. The agreement governs how listings may be displayed, how long data may be retained, what must be attributed, and whether anything may pass downstream. National coverage therefore means a separate relationship with each of hundreds of organisations, which is precisely the work commercial aggregators are selling.
RESO standards help with the shape of the data, not the right to it. A common field schema and a modern web API mean your integration code is reusable across MLSs that adopt them, while the legal and commercial work stays per-MLS. Teams routinely underestimate the second half and budget only for the first.
What can you build without MLS access?
| Product | Use this source when | Avoid when |
|---|---|---|
| Valuation and market analytics | Public records: transfers, assessments, parcels | You need live listing status, which records do not carry |
| Neighbourhood and context data | Open government sources: permits, zoning, schools, transit | Never scrape what is published as open data |
| Live listings in a consumer product | A licensed feed, via membership or an aggregator | You planned to take them from portal pages |
| Rental market analysis | Sources whose terms permit it, checked individually | The platform prohibits automated collection |
Two practical notes on the open half. County systems are small and often slow, so collection should be gentle and spread out, and some restrict access from outside their region, which shows up as thin coverage rather than an error. DataImpulse residential gives a country or state-pinned exit at $1 per GB when that is what a portal expects.
What are the limits?
Portal terms are binding regardless of public visibility. The major property portals prohibit automated extraction explicitly, and their estimates are proprietary products. A page being loadable in a browser does not change the contract you accept by using the service.
Retention and display rules travel with licensed data. MLS agreements commonly limit how long records may be kept and require specific attribution, and those obligations do not disappear when the data enters your warehouse.
Personal data appears in public records. Owner names and addresses are public in most US counties and still personal data under several privacy regimes, particularly if you serve users outside the US.
Coverage is county by county. Any national claim is an aggregate over jurisdictions with very different digitisation, so measure coverage in the markets you actually serve. General information, not legal advice.
Related: public records access, is web scraping legal.
Frequently Asked Questions
Is there a real estate data API I can just buy?
For listings, not directly. MLS feeds are granted to licensed participants under per-organisation agreements rather than sold publicly, so commercial access usually means an aggregator that has already done that work across many regional MLSs.
What does RESO standardise?
The data shape: a common field dictionary and a modern web API, so integration code is reusable across MLSs that adopt them. It does not standardise the right to access, which remains a separate agreement with each MLS.
Can I scrape property listings from portals?
Their terms almost universally prohibit it, and that is a contractual matter independent of the page being publicly viewable. Products that need live listings are built on licensed feeds; products built on public records avoid the question entirely.
What can I build from public records alone?
Valuation models, market analytics, ownership research and neighbourhood context, using county transfers, assessments, parcels and permits. What public records do not carry is live listing status, which is the specific thing MLS access provides.
Why does coverage vary so much by area?
Because county systems digitise at very different rates and MLSs operate independently. A national coverage figure is an average across jurisdictions that range from modern APIs to paper-derived exports, so measure coverage in the markets you serve.
Reach county systems the way they expect
County and state portals sometimes restrict access by origin, and thin coverage looks like a gap rather than a block. DataImpulse residential proxies give country and state-level exits at $1 per GB across 195 countries. Create an account and check one portal from its own region.
Related: public court records · government contract data · is web scraping legal.
Last updated: September 17, 2026.
State/City/Zip/ASN Targeting 
