The finding
The percentage of homes with flood insurance through the National Flood Insurance Program is 2.89% nationally: 3,160,864 residential policies in force against an estimated 109,326,879 residential structures, according to FEMA figures as of 3 August 2026. That is roughly one home in 35. Inside the mapped Special Flood Hazard Area the rate is 43.9%; outside it, where 96.1% of homes sit, it is 1.25%. The typical county is far lower than the national figure: its median is 0.44%.
Headline figures
In this report
- Why the national rate misleads
- Every state, ranked
- Inside and outside the mapped flood zone
- Where the policies are
- County extremes
- Declared again and again, rarely covered
- What the file gets wrong, and how it was handled
- Method
- Reproduce this
- What this does not show
- Questions this report answers
- Sources
Key findings
- FEMA counts 3,160,864 NFIP residential policies against 109,326,879 residential structures, a national take-up of 2.89%. About 106.2 million homes have no NFIP policy.
- The median county sits at 0.44%, and 2,379 of 3,144 counties (75.7%) are under 1%. The national rate is pulled up by a small set of coastal counties.
- Inside the Special Flood Hazard Area, take-up is 43.9%; outside it, 1.25%, a gap of 35 times. Mandatory purchase rules apply only inside the zone.
- 96.1% of residential structures sit outside the mapped zone, so 97.6% of the homes without an NFIP policy are outside it: 93.0 million, against 2.26 million inside.
- The 100 counties with the most policies hold 73.9% of all NFIP residential policies; the top 10 hold 30.5%. Half of all policies sit in 29 counties.
- Louisiana leads at 22.0%, followed by Florida at 13.7%. Minnesota is lowest at 0.23%, a spread of 97 times.
- 5 counties with at least 10,000 residential structures have zero NFIP residential policies in force.
- Counties with 10 or more storm and flood major disaster declarations since 2015 have a median take-up of 1.01%, and 49% of them are under 1%, against a median of 0.40% in counties with none. Repeated declarations go with only modestly higher NFIP cover.
Why the national rate misleads
A national rate of 2.89% sounds like a figure that describes the country. It describes almost no county. FEMA’s file has 3,144 counties with a structure estimate, and the median county rate is 0.44%; the pooled national figure is 6.5 times the median. The ninetieth percentile county is at 2.42%, still below the national rate. Only 270 counties (8.6%) sit above it.
2,379 counties, 75.7% of the total, are under 1%. 194 record no NFIP residential policy at all. The pooled rate is weighted toward the counties with the most homes and the most policies, and those are overwhelmingly on the Gulf and Atlantic coasts. For a restoration firm working an inland flood, the median county rate is the more realistic prior for how many affected homes carry NFIP cover; the national rate overstates it.
Both numbers are correct; they answer different questions. The national rate answers what share of all US homes carry an NFIP policy. The county median answers what take-up looks like in a typical place. Journalists quoting either should name which one it is. The claim side of the same picture is in our flood claims by state report.
Every state, ranked
| State | Policies in force | Residential structures | Take-up | Median county | Inside mapped zone |
|---|---|---|---|---|---|
| Louisiana | 359,013 | 1,629,418 | 22.03% | 7.27% | 56.9% |
| Florida | 1,040,464 | 7,585,072 | 13.72% | 5.81% | 66.9% |
| South Carolina | 121,413 | 1,976,591 | 6.14% | 0.49% | 73.1% |
| Texas | 504,754 | 8,773,166 | 5.75% | 0.61% | 40.6% |
| New Jersey | 121,462 | 2,419,797 | 5.02% | 1.91% | 62.4% |
| Delaware | 17,547 | 366,216 | 4.79% | 1.27% | 58.1% |
| Hawaii | 11,829 | 304,141 | 3.89% | 4.10% | 54.0% |
| Mississippi | 40,942 | 1,163,888 | 3.52% | 0.59% | 32.7% |
| Rhode Island | 8,360 | 306,168 | 2.73% | 4.88% | 49.0% |
| Virginia | 72,514 | 2,733,160 | 2.65% | 0.68% | 28.9% |
| North Carolina | 106,560 | 4,056,295 | 2.63% | 0.81% | 53.3% |
| New York | 118,000 | 4,700,087 | 2.51% | 0.89% | 55.5% |
| Connecticut | 21,002 | 997,618 | 2.11% | 2.64% | 51.0% |
| Massachusetts | 36,939 | 1,819,669 | 2.03% | 1.70% | 49.4% |
| North Dakota | 4,925 | 276,636 | 1.78% | 0.24% | 26.8% |
| Maryland | 33,232 | 1,925,531 | 1.73% | 1.16% | 64.1% |
| Georgia | 58,557 | 3,659,988 | 1.60% | 0.32% | 42.9% |
| New Mexico | 11,342 | 807,285 | 1.40% | 0.58% | 14.9% |
| California | 140,683 | 10,537,796 | 1.34% | 1.02% | 30.3% |
| Alabama | 25,579 | 1,923,527 | 1.33% | 0.36% | 29.4% |
| Maine | 6,651 | 538,014 | 1.24% | 0.74% | 22.1% |
| District of Columbia | 1,443 | 125,848 | 1.15% | 1.15% | No estimate |
| Alaska | 2,707 | 242,920 | 1.11% | 0.42% | 33.3% |
| Vermont | 2,848 | 257,879 | 1.10% | 0.89% | 27.1% |
| New Hampshire | 4,490 | 414,294 | 1.08% | 0.77% | 25.2% |
| Oregon | 16,155 | 1,497,491 | 1.08% | 0.93% | 23.8% |
| West Virginia | 7,746 | 763,535 | 1.01% | 0.87% | 10.3% |
| Washington | 23,122 | 2,481,708 | 0.93% | 0.77% | 28.7% |
| Kentucky | 14,136 | 1,733,651 | 0.82% | 0.43% | 18.8% |
| Nevada | 7,964 | 1,038,659 | 0.77% | 1.20% | 29.6% |
| Arkansas | 8,993 | 1,189,681 | 0.76% | 0.53% | 14.0% |
| Tennessee | 18,014 | 2,579,192 | 0.70% | 0.38% | 24.5% |
| Pennsylvania | 31,965 | 4,599,973 | 0.69% | 0.67% | 22.0% |
| Nebraska | 5,587 | 837,251 | 0.67% | 0.19% | 14.4% |
| Arizona | 17,476 | 2,647,475 | 0.66% | 0.68% | 21.3% |
| Montana | 3,070 | 523,618 | 0.59% | 0.33% | 14.3% |
| Colorado | 11,282 | 1,993,984 | 0.57% | 0.34% | 21.1% |
| Idaho | 3,768 | 691,150 | 0.55% | 0.27% | 15.0% |
| Illinois | 21,080 | 3,952,182 | 0.53% | 0.32% | 22.0% |
| Indiana | 13,093 | 2,572,454 | 0.51% | 0.37% | 16.5% |
| South Dakota | 2,011 | 398,954 | 0.50% | 0.12% | 13.2% |
| Wyoming | 1,305 | 272,148 | 0.48% | 0.27% | 10.5% |
| Iowa | 6,602 | 1,396,605 | 0.47% | 0.22% | 12.6% |
| Oklahoma | 6,994 | 1,514,753 | 0.46% | 0.25% | 12.3% |
| Wisconsin | 8,929 | 1,939,288 | 0.46% | 0.36% | 18.4% |
| Kansas | 5,331 | 1,172,094 | 0.45% | 0.20% | 14.0% |
| Missouri | 10,574 | 2,496,627 | 0.42% | 0.19% | 15.0% |
| Ohio | 18,040 | 4,322,423 | 0.42% | 0.40% | 15.7% |
| Michigan | 16,036 | 3,919,085 | 0.41% | 0.25% | 16.2% |
| Utah | 3,169 | 979,402 | 0.32% | 0.30% | 15.2% |
| Minnesota | 5,166 | 2,272,492 | 0.23% | 0.23% | 11.6% |
Louisiana (22.0%) and Florida (13.7%) are in a category of their own; the next state, South Carolina, is at 6.1%. At the other end, 10 states have fewer than one NFIP policy for every 200 homes (under 0.5%): Minnesota, Utah, Michigan, Ohio, Missouri, Kansas, Wisconsin, Oklahoma, Iowa, Wyoming.
Compare the take-up column with the median county column for any coastal state and the same skew appears at state scale. A state rate is an average weighted toward its largest coastal counties; its interior counties usually look like the bottom of this table.
Inside and outside the mapped flood zone
The Special Flood Hazard Area is the zone FEMA maps as having a 1% or greater annual chance of flooding. Federally backed mortgages on buildings inside it carry a flood insurance purchase requirement, and take-up there is of a different order. Across the 2,339 counties where FEMA estimates a non-zero number of homes inside the zone, take-up there is 43.9%. Outside it is 1.25%.
The quotable line is the one in the bottom row. Because 96.1% of residential structures sit outside the mapped zone, the zone accounts for only 2.4% of the homes without an NFIP policy. 93.0 million homes without a policy are outside the zone, against 2.26 million inside. Policies outside the zone are not trivial either: they make up 41.1% of all policies in these counties.
For restoration demand this matters because flood damage does not stop at the mapped line. Our flood zone exposure report sets out how much NFIP claim activity comes from outside the high-risk zones. Outside the zone, about 1.2 homes in 100 carry an NFIP policy on these figures, so when a home there floods, the cost of the job far more often falls on the owner or on capped public assistance, as set out in our disaster grant ledger. This report describes the data and is not advice on whether any property should carry cover; FloodSmart and state insurance departments are the places for that decision.
Where the policies are
NFIP residential cover is concentrated to a degree that the state table hides. Of 3,160,864 policies, 30.5% are in just ten counties and 73.9% in 100. Half of all policies are in 29 of 3,144 counties.
| County | Policies in force | Residential structures | Take-up | Share of national policies |
|---|---|---|---|---|
| Harris, Texas | 206,279 | 1,245,880 | 16.6% | 6.53% |
| Miami-Dade, Florida | 164,154 | 492,181 | 33.4% | 5.19% |
| Broward, Florida | 125,204 | 652,397 | 19.2% | 3.96% |
| Lee, Florida | 84,120 | 346,951 | 24.2% | 2.66% |
| Pinellas, Florida | 76,020 | 391,677 | 19.4% | 2.41% |
| Jefferson, Louisiana | 75,564 | 116,441 | 64.9% | 2.39% |
| Palm Beach, Florida | 69,154 | 530,976 | 13.0% | 2.19% |
| Orleans, Louisiana | 56,109 | 108,361 | 51.8% | 1.78% |
| Hillsborough, Florida | 54,179 | 448,186 | 12.1% | 1.71% |
| Galveston, Texas | 52,889 | 125,276 | 42.2% | 1.67% |
| Collier, Florida | 51,467 | 174,174 | 29.5% | 1.63% |
| Charleston, South Carolina | 46,856 | 148,009 | 31.7% | 1.48% |
| Sacramento, California | 42,334 | 443,240 | 9.6% | 1.34% |
| Fort Bend, Texas | 39,537 | 255,919 | 15.4% | 1.25% |
| Sarasota, Florida | 37,772 | 210,450 | 17.9% | 1.19% |
Harris, Texas has the most policies, 206,279, but its take-up is 16.6% because it also has 1,245,880 homes. Volume and rate are different stories: the county where the most policies sit is not the county where the largest share of homes is covered.
County extremes
County rates are only meaningful above a minimum size, because a county with a few hundred homes can swing on a handful of policies. The tables below use the 1,760 counties with at least 10,000 estimated residential structures.
| County | Policies in force | Residential structures | Take-up | Inside mapped zone |
|---|---|---|---|---|
| Jefferson, Louisiana | 75,564 | 116,441 | 64.9% | 113.1% |
| Monroe, Florida | 22,181 | 38,080 | 58.2% | 61.3% |
| St. Bernard, Louisiana | 8,351 | 14,712 | 56.8% | 113.7% |
| St. Charles, Louisiana | 9,907 | 17,783 | 55.7% | No estimate |
| Orleans, Louisiana | 56,109 | 108,361 | 51.8% | 115.2% |
| Galveston, Texas | 52,889 | 125,276 | 42.2% | 66.0% |
| St. Tammany, Louisiana | 37,387 | 92,439 | 40.4% | No estimate |
| Cape May, New Jersey | 25,975 | 64,831 | 40.1% | 74.4% |
| Dare, North Carolina | 12,409 | 31,280 | 39.7% | 84.0% |
| Beaufort, South Carolina | 28,198 | 71,094 | 39.7% | 102.0% |
| St. John the Baptist, Louisiana | 5,448 | 15,943 | 34.2% | 79.8% |
| Miami-Dade, Florida | 164,154 | 492,181 | 33.4% | 74.5% |
| Hancock, Mississippi | 6,730 | 20,309 | 33.1% | 63.8% |
| Terrebonne, Louisiana | 10,803 | 33,598 | 32.2% | No estimate |
| Charlotte, Florida | 32,247 | 100,666 | 32.0% | 57.9% |
Jefferson, Louisiana leads at 64.9%, followed by Monroe, Florida at 58.2%. Even at the top of the county list, 35.1% of homes in Jefferson, Louisiana have no NFIP policy.
At the bottom, 5 counties of this size record zero NFIP residential policies, all in Michigan and Minnesota, and 91 are under 0.1%. FEMA’s file estimates 0 homes inside the mapped flood zone across the zero-policy counties, though that estimate is missing for many counties (see the data quality section). The file does not say whether each county’s communities participate in the NFIP, so a zero should not be read as a decision by owners not to buy.
Declared again and again, rarely covered
Do the counties that keep getting declared disasters carry more flood cover? To test it, county take-up was joined to FEMA major disaster declarations for flood, hurricane, severe storm, tropical storm, coastal storm, typhoon and dam or levee events declared from 2015 to 2026. County keys were built from the state FIPS code and FEMA’s county place code. 2,526 of 2,613 declared county keys matched the take-up file, a match rate of 96.7%. The 87 unmatched keys break down as Puerto Rico 78, Connecticut 8, South Dakota 1. Puerto Rico is not in the take-up file, and Connecticut declarations use the state’s newer planning region codes rather than its former county codes.
| Distinct declarations | Counties | Pooled take-up | Median county take-up | Counties under 1% |
|---|---|---|---|---|
| None | 618 | 1.04% | 0.40% | 82.0% |
| 1 to 2 | 1,085 | 0.95% | 0.39% | 81.8% |
| 3 to 5 | 1,078 | 2.85% | 0.44% | 75.1% |
| 6 to 9 | 322 | 10.92% | 1.11% | 47.8% |
| 10 or more | 41 | 3.43% | 1.01% | 48.8% |
The pattern is not a steady climb. The pooled rate peaks in the 6 to 9 band at 10.92%, the band that holds the large hurricane-declared coastal counties (its three biggest by policy count are Harris, Texas; Broward, Florida; Lee, Florida), and falls back to 3.43% in the 10 or more band. The median county moves less: from 0.40% in counties with no such declaration to 1.01% in counties with 10 or more, and 48.8% of that most-declared group are still under 1%.
| County | Distinct declarations | Take-up | Residential structures |
|---|---|---|---|
| Owsley, Kentucky | 17 | 0.30% | 2,016 |
| Clay, Kentucky | 16 | 0.68% | 7,843 |
| Johnson, Kentucky | 14 | 2.31% | 9,151 |
| Lee, Kentucky | 14 | 0.57% | 2,960 |
| Breathitt, Kentucky | 13 | 1.51% | 7,764 |
| Lawrence, Kentucky | 13 | 0.42% | 7,597 |
| Metcalfe, Kentucky | 13 | 0.23% | 4,417 |
| Whitley, Kentucky | 13 | 0.28% | 14,201 |
| Estill, Kentucky | 12 | 0.74% | 6,101 |
| Floyd, Kentucky | 12 | 3.35% | 17,217 |
The most declared counties in this period are in Kentucky, and none of the ten has take-up above 3.4%. The median of the ten has 7,680 residential structures: small housing stocks where declaration follows declaration while NFIP take-up stays in low single digits or below. The full declaration ranking is in our most declared counties report, and the properties that flood repeatedly are tracked in repetitive flood loss properties. Nothing here establishes why take-up is low in these counties; income, mapping, mortgage status and community participation are all outside this file.
What the file gets wrong, and how it was handled
Three problems in the FEMA file change the numbers if they are ignored. First, 15 rows, mostly labelled Unknown Counties, have a structure estimate of zero but carry 372 policies between them. They have no denominator, so they are left out of every rate. No county has total policies above its total structure count.
Second, 593 counties report policies inside the Special Flood Hazard Area but a zero estimate of homes inside it, accounting for 152,001 SFHA policies. The largest are St. Tammany, Louisiana (17,097), Nueces, Texas (10,928), Terrebonne, Louisiana (7,223). Dividing national SFHA policies by national SFHA structures without removing these gives 47.9%, a figure that puts policies on top with no homes beneath them. This report uses 43.9%, computed only over counties with a non-zero SFHA structure estimate.
Third, 27 of those counties still show more SFHA policies than SFHA homes, for example Pinellas, Florida (56,785 policies against 4,807 structures), Jefferson, Louisiana (41,511 policies against 36,689 structures), Collier, Florida (38,559 policies against 152 structures). A policy count is a count of contracts, and condominium buildings can carry one contract per unit while counting as one structure, which is one plausible contributor; the file does not separate them. These counties are kept, so the zone rate is a slight overstatement rather than a cleaned estimate.
Method
How we counted. We retrieved 3,159 rows from FEMA OpenFEMA NfipResidentialPenetrationRates v1 on 5 October 2026; every row carries the as-of date 3 August 2026. Take-up is resContractsInForce divided by totalResStructures. National and state rates are pooled (summed policies over summed structures), never averages of county rates. 15 rows with a zero structure estimate, holding 372 of 3,161,236 policies, are excluded from all rates, leaving 3,144 counties. The FEMA-published rate field was checked against the computed ratio and agrees to rounding where present; it is null on some rows, so the ratio is computed directly. The SFHA split uses the 2,339 counties with a non-zero totalResStructuresSfha; outside the zone is the county total minus the SFHA figure. Homes without an NFIP policy are counted per county as structures minus policies, floored at zero, then summed. County extremes use a floor of 10,000 residential structures. The declaration cross-reference uses 8,943 major disaster (DR) designated-area rows for storm and flood incident types declared 2015 to 2026; 278 rows naming tribal areas or other non-county places were dropped, and a county counts each disaster number once.
Reproduce this
Reproduce this
- Endpoint
https://www.fema.gov/api/open/v1/NfipResidentialPenetrationRates- Query
$top=10000 (single page; 3,159 rows returned)- Rows retrieved
- 3,159 county rows
- Retrieved
- 5 October 2026
No API key is required. Every value is serialised as a number or null; the rate fields are fractions, not percentages. For the declaration cross-reference, pull https://www.fema.gov/api/open/v2/DisasterDeclarationsSummaries with $filter=declarationDate ge '2015-01-01', keep declarationType DR, and build the county key as the two-digit state FIPS code followed by the last three digits of placeCode where placeCode starts with 99. That join matched 96.7% of declared county keys.
What this does not show
What this does not show
- NFIP only. Private flood insurance is not in the FEMA file. A home counted here as having no NFIP policy may carry private flood cover, so the true share of homes with any flood insurance is somewhat higher than these figures.
- Structure counts are FEMA estimates. The denominator is an estimated count of residential structures, not a census of homes or households, and the SFHA estimate is missing for many counties.
- One date. Policies in force on 3 August 2026. Policies lapse and are bought seasonally, so take-up on another date will differ, and this file gives no trend.
- Residential only. Commercial and other non-residential NFIP policies are not counted.
- Contracts, not homes. A policy is a contract. Condominium units and buildings can be counted differently from structures, which is why some SFHA ratios exceed 100%.
- No causes. The file cannot say why take-up is high or low. Mortgage requirements, community participation, premiums, income and map changes are all outside it.
- Declarations are not floods. The cross-reference counts storm and flood major disaster declarations naming a county, not measured flood events or damage.
Most quotable figures
- Only 2.89% of US residential structures carry an NFIP flood insurance policy: 3,160,864 policies against 109,326,879 structures (FEMA OpenFEMA, as of 3 August 2026).
- The median US county has an NFIP residential take-up rate of 0.44%, and 2,379 of 3,144 counties are under 1% (FEMA OpenFEMA, as of 3 August 2026).
- 93.0 million homes without an NFIP policy sit outside FEMA’s mapped flood zone, 97.6% of all homes without one (FEMA OpenFEMA, as of 3 August 2026).
- Inside FEMA’s mapped flood zone take-up is 43.9%; outside it is 1.25% (FEMA OpenFEMA, as of 3 August 2026).
- 100 counties hold 73.9% of all NFIP residential policies in the United States (FEMA OpenFEMA, as of 3 August 2026).
- Louisiana has the highest NFIP residential take-up at 22.0%; Minnesota the lowest at 0.23% (FEMA OpenFEMA, as of 3 August 2026).
Questions this report answers
What percentage of homes have flood insurance?
Through the NFIP, 2.89% of US residential structures as of 3 August 2026: 3,160,864 policies against 109,326,879 estimated structures. Private flood policies are not included, so the share with any flood cover is somewhat higher.
What share of homes in a flood zone have flood insurance?
Inside the Special Flood Hazard Area, NFIP take-up is 43.9% across the 2,339 counties where FEMA estimates homes in the zone. Outside it, take-up is 1.25%.
Which state has the most homes with flood insurance?
By share of homes, Louisiana at 22.0%, then Florida at 13.7% and South Carolina at 6.1%. The lowest is Minnesota at 0.23%.
Why is the national rate higher than most counties?
Because it is pooled. Coastal counties with many homes and high take-up dominate the total, while the median county is at 0.44%.
Should I buy flood insurance?
This report describes FEMA data and does not give insurance advice. FloodSmart.gov, a property’s flood map on FEMA’s Map Service Center and the state insurance department are the places to start that decision.
Can I reuse these figures?
Yes. The tables and derived figures are released under CC BY 4.0 with attribution to Remediation Insider. The FEMA data is public and reproducible from the endpoint above.
How to cite
Remediation Insider. Percentage of Homes With Flood Insurance: The County Take-Up Map. https://remediationinsider.com/flood-insurance-take-up-map/ (retrieved 5 October 2026).
Data and tables in this report are released under CC BY 4.0. Attribution: Remediation Insider.
Sources
- Federal Emergency Management Agency. FEMA OpenFEMA NFIP Residential Penetration Rates, v1. 3,159 county rows, as of 3 August 2026, retrieved 5 October 2026.
- Federal Emergency Management Agency. FEMA OpenFEMA Disaster Declarations Summaries, v2. Major disaster designated-area rows declared from 2015 onward.
- FEMA flood maps and the Special Flood Hazard Area.
- FloodSmart, the official NFIP site, for coverage questions.
- FEMA OpenFEMA Terms and Conditions.
- No survey, interview or physical inspection was conducted for this report. Every figure is computed at build time from the retrieved FEMA files.
This product uses the Federal Emergency Management Agency’s OpenFEMA API, but is not endorsed by FEMA. The Federal Government or FEMA cannot vouch for the data or analyses derived from these data after the data have been retrieved from the Agency’s website(s).