Florida Sues First Street Over Flood Scores. Who Is Right About Risk?

Florida says First Street's property-level flood scores mislead buyers. The bigger question is how much confidence consumers should place in a risk score when models disagree.

A score beside a home listing can now land in a courtroom. On Oct. 23, Florida Attorney General James Uthmeier sued First Street Technology in Polk County, alleging its Risk Factor scores exaggerate flood danger, are hard to challenge and may harm sellers. The case raises a broader question: What does a 1-to-10 score tell buyers, lenders and insurers, and how much confidence does it deserve? Florida says the numbers mislead. FEMA maps miss some risk, too; neither source settles the question.

What the scores are

First Street assigns 1-to-10 Risk Factor ratings for flood, wildfire, wind, heat and air quality. Its Flood Factor methodology combines 30-year flood likelihood with projected depth, covering rainfall, rivers, tides, storm surge and future conditions. A 1% annual chance equals about 26% over 30 years. Flood Factor is not an annual probability; depth also affects the score.

The scores have appeared on Zillow, Redfin and Realtor.com; the complaint says Zillow removed direct display in late 2025. It also says Fannie Mae, Freddie Mac and insurers use First Street data in property assessments. First Street says its property-level model captures heavy rainfall and future conditions that FEMA maps may miss. MSCI completed its acquisition of First Street on August 3, 2026. Different methods can produce different scores, but that difference alone does not establish accuracy at a particular address.

What Florida alleges

The Florida complaint, filed under the state's Deceptive and Unfair Trade Practices Act, makes these allegations, which have not been tested in court:

• A score differs from FEMA's map. An Oldsmar home in Zone X, labeled minimal flood hazard, received a First Street score of 7, or "Major."
• Florida alleges a score-method mismatch. The complaint pairs that 7 with a 74%–82% 30-year flood probability. It quotes First Street's methodology saying properties with at least a 99% chance score 6 or higher. That is a one-way statement; it does not establish that every score of 6+ means a 99% chance. The complaint's reading of the method has not been tested in court.
• Florida questions other city-level ratings. The complaint says First Street listed coastal flooding for Orlando, an inland city, and rated Lake City's air risk "Major" despite the EPA saying the city met pollution standards.
• Florida says disclosures and corrections fall short. Florida challenges claims that scores are peer-reviewed and accurate at property level, says underlying data are not readily available, and cites First Street guidance that individual disputes generally do not change Factor scores.

Florida cites Zillow's analysis of U.S. homes listed in June 2024: 52% with extreme flood scores (9 or 10) sold, versus 71% with low scores (1 or 2). Zillow says the gap has persisted since 2018 and widened. This association does not show the scores caused the difference. Florida also alleges impacts on insurance costs and coverage, but the complaint does not establish a pricing effect.

Florida seeks an injunction, consumer redress, disgorgement and civil penalties of $10,000 per violation, or $15,000 for violations involving a senior or person with a disability. Nebraska's attorney general filed a similar case in August. Both cases remain allegations. On Oct. 27, Florida Phoenix/WLRN reported it had sought comment from First Street but had not heard back.

Do the scores overstate risk?

Research shows uncertainty in property-level models, especially in cities, but not whether First Street overstates risk.

A 2024 Earth's Future study by UC Irvine and University of Miami researchers compared First Street with PRIMo and drainage-resolving PRIMo-Drain in Los Angeles. For the 1%-annual-chance flood extent, the county-level agreement index was 24% between First Street and PRIMo, and 20% with PRIMo-Drain. These results are specific to Los Angeles, not proof that 76% of First Street scores nationwide are wrong. The authors warn national models may not reliably pinpoint urban properties at risk.

A 2025 GARP study benchmarked 13 physical-risk vendors on 100 properties and found substantial variation in hazard and damage estimates. That signals an industry-wide challenge, not proof against First Street alone.

FEMA maps help set floodplain rules and determine whether a federally backed mortgage requires flood insurance. FEMA's Risk Rating 2.0, introduced in 2021, prices NFIP coverage using property-specific data, multiple flood types and catastrophe models. A Zone X label is not a current premium estimate or a promise a home will not flood.

The National Flood Insurance Program says nearly one-third of its 2014–2024 claims came from outside high-risk areas. That underscores the limits of zone labels; property-level scores still need to be evaluated against their own assumptions.

What it means for insurance

A property-level score is a model output; it does not replace a site inspection. Buyers may care about a 30-year possibility, while insurers need a defensible view of expected loss and uncertainty.

For insurance professionals:

1. Treat the score as one input. Check the peril, time horizon, depth assumptions and local drainage or mitigation.
2. Don't read Zone X as "safe." Explain the limits of both the map and the model.
3. Ask vendors to show their work. Clear definitions, uncertainty ranges and a usable dispute process matter when scores influence purchases and coverage.

Florida's case puts disclosure and correction practices before a court. Whether a particular First Street score overstates risk is an address-by-address question; users need the methods, assumptions and uncertainty behind it.

Sources: Florida complaint, Florida Phoenix/WLRN, First Street flood methodology, Zillow Research, UC Irvine study release, GARP 2025 study, FEMA Risk Rating 2.0 FAQ, FEMA FloodSmart claims data, MSCI acquisition release, and Nebraska complaint. The Florida complaint's assertions remain allegations.