At least 18,000,208 American homes, 15.7% of every residential address in the country, have at most one wired internet company able to serve them.
That is a floor rather than an estimate. The count comes from what the providers themselves filed with the FCC, and providers overstate their coverage rather than understate it; every overstatement moves an address into this group and none move out. Measured block by block across all 6,280,236 US census blocks in filing period D25, published July 21, 2026.
Free to cite and to reproduce with attribution to WhereFiber. Method, the two directions the data is biased in, and the figure for every state are all below. Computed by Ferenc Horvath, independently and with no provider funding. Questions: contact@wherefiber.com.
Every finding
| Finding | Share of US residential locations |
|---|---|
| Have at most one wireline company reporting service | 15.7% |
| Have no wireline company reporting service at all | 1.5% |
| Have at most one company reporting fiber | 80.1% |
| Are in a block where nobody reports fiber | 27.9% |
| Are in a block where two or more companies report fiber | 19.9% |
The number worth sitting with is the third one. Roughly four in five US residential locations have at most one fiber option, and more than a quarter have none. For most of the country, “which fiber provider should I pick” is not a question that arises.
The gap nobody reports: fiber that exists but does not reach you
Coverage statistics are usually published at state or county level, which hides the thing that actually decides your options. Fiber was built out street by street, so a block can contain fiber and still not offer it to every building in that block.
| Where a location stands on fiber | Share |
|---|---|
| Reached by at least one fiber operator | 60.4% |
| In a block that has fiber, but not reached by it | 11.8% |
| In a block with no fiber presence at all | 27.9% |
That middle row is up to one in nine American homes, and up to is doing real work in that sentence. The figure is what is left after subtracting our fiber-reach floor, and that floor counts only the largest single fiber operator in each block. Where two operators cover different halves of one block, their combined reach is understated and this residual is correspondingly inflated. Treat it as a ceiling. The opposing bias, providers overstating coverage, pushes it back down by an amount we cannot measure, which is exactly why we publish both directions instead of a single confident figure. Those homes appear as “fiber available” on every map published at county level or above, and cannot buy it. It is the most common reason a coverage map disagrees with a provider's own address check.
How concentrated is fiber?
Across the 54 operators we have mapped to consumer brands, the top five account for 62% of reported residential fiber locations, and the top ten for 77%.
| # | Operator | Residential fiber locations reported |
|---|---|---|
| 1 | AT&T Fiber | 19,375,365 |
| 2 | Verizon Fios | 9,364,253 |
| 3 | Frontier Fiber | 6,545,956 |
| 4 | Quantum Fiber | 3,157,734 |
| 5 | Metronet | 2,612,684 |
| 6 | Spectrum | 2,284,620 |
| 7 | Optimum | 2,277,657 |
| 8 | Google Fiber | 1,886,259 |
| 9 | Brightspeed | 1,793,486 |
| 10 | Kinetic | 1,443,622 |
Counted across operators we have mapped from FCC holding-company names to the brand consumers actually buy, which is not a mapping the FCC publishes. Smaller unmapped operators are excluded, so the true long tail is longer than this table implies and the concentration figures are, if anything, slightly overstated.
Every state ranked
Rank 1 is the state where the fewest homes are stuck with one wired company or none; rank 52 is where the most are. The percentage is the share of residential locations in a census block where at most one company reports wireline service, whether fiber, cable or DSL. Sorted worst first.
| Rank | State | Locations | At most one wireline company | No fiber in the block |
|---|---|---|---|---|
| 52 | North Dakota | 324,994 | 39.3% | 22.2% |
| 51 | Massachusetts | 1,945,048 | 37.9% | 38.5% |
| 50 | Alaska | 257,108 | 36.6% | 69.4% |
| 49 | South Dakota | 391,050 | 35.8% | 17.5% |
| 48 | Delaware | 404,117 | 35.5% | 33.5% |
| 47 | Montana | 467,248 | 33.9% | 46.9% |
| 46 | Wyoming | 251,272 | 28.2% | 41.1% |
| 45 | Maryland | 1,849,936 | 26.8% | 25.4% |
| 44 | New Jersey | 2,567,038 | 25.6% | 23.8% |
| 43 | Pennsylvania | 4,810,145 | 25.2% | 30.3% |
| 42 | Idaho | 748,192 | 24.6% | 43.1% |
| 41 | New York | 4,594,598 | 24.2% | 19.3% |
| 40 | Oklahoma | 1,632,515 | 23.3% | 20.4% |
| 39 | Virginia | 2,838,226 | 23.2% | 19.8% |
| 38 | Kansas | 1,179,222 | 23.1% | 15.9% |
| 37 | Iowa | 1,348,996 | 22.8% | 16.0% |
| 36 | West Virginia | 807,459 | 20.8% | 37.0% |
| 35 | Nebraska | 784,438 | 20.7% | 16.3% |
| 34 | Kentucky | 1,799,563 | 20.6% | 19.3% |
| 33 | New Mexico | 841,825 | 20.1% | 62.7% |
| 32 | Missouri | 2,486,243 | 19.7% | 24.3% |
| 31 | Wisconsin | 2,261,131 | 18.8% | 34.6% |
| 30 | Mississippi | 1,286,102 | 18.4% | 20.8% |
| 29 | Arkansas | 1,310,979 | 18.3% | 23.6% |
| 28 | Michigan | 3,934,168 | 17.2% | 49.2% |
| 27 | Alabama | 2,111,762 | 15.5% | 23.4% |
| 26 | Louisiana | 1,778,525 | 15.4% | 27.7% |
| 25 | Minnesota | 2,099,900 | 15.2% | 28.5% |
| 24 | South Carolina | 2,205,891 | 14.8% | 21.2% |
| 23 | Texas | 10,104,973 | 14.5% | 16.4% |
| 22 | Arizona | 2,725,868 | 14.4% | 52.3% |
| 21 | Washington | 2,520,667 | 13.7% | 47.5% |
| 20 | Illinois | 4,091,096 | 13.3% | 48.1% |
| 19 | Indiana | 2,700,879 | 12.1% | 21.6% |
| 18 | Utah | 1,024,100 | 11.9% | 14.8% |
| 17 | Vermont | 279,292 | 11.8% | 20.7% |
| 16 | Rhode Island | 335,753 | 11.0% | 7.0% |
| 15 | Ohio | 4,481,610 | 10.8% | 31.1% |
| 14 | Colorado | 1,971,093 | 10.5% | 36.0% |
| 13 | Tennessee | 2,761,872 | 10.3% | 13.3% |
| 12 | California | 10,112,606 | 10.1% | 39.1% |
| 11 | Georgia | 3,820,124 | 9.8% | 16.3% |
| 10 | Nevada | 1,020,073 | 9.5% | 44.4% |
| 9 | Oregon | 1,487,199 | 9.4% | 28.5% |
| 8 | North Carolina | 4,310,745 | 8.8% | 20.6% |
| 7 | Maine | 617,854 | 8.6% | 22.5% |
| 6 | Puerto Rico | 1,170,206 | 7.9% | 24.7% |
| 5 | Florida | 7,576,700 | 7.6% | 21.5% |
| 4 | New Hampshire | 513,599 | 4.5% | 12.8% |
| 3 | Hawaii | 310,022 | 4.3% | 4.0% |
| 2 | District of Columbia | 123,361 | 1.8% | 4.1% |
| 1 | Connecticut | 1,084,189 | 1.4% | 2.8% |
States and territories with fewer than 50,000 reported residential locations are omitted.
Method, and what these numbers cannot tell you
Source: the FCC Broadband Data Collection, filing period D25, published July 21, 2026. We processed the location-level availability files into a per-census-block index (6,280,236 blocks, 114,550,235 residential locations) and counted distinct companies by FCC provider ID, so AT&T Fiber and AT&T Internet count as one company rather than two.
Two biases pull in opposite directions. Both are stated here.
This understates how many homes have one option. The data records how many locations in a block a company can serve, not which ones. A block where two companies report is not a block where every building has two choices, but some will still have one. Counting only the one-company blocks therefore produces a floor, not an estimate. The real figure is higher than 15.7%.
And this overstates it, in the other direction. The unit here is a residential location, meaning an address or building, not a household. A fifty-unit apartment building counts once, the same as a single-family home. Multi-unit buildings skew urban and urban areas are more competitive, so a household-weighted version of these figures would show more choice than a location-weighted one.
We have not tried to net those two effects against each other, because we cannot do so honestly from this data. They are stated so you can judge the direction yourself.
Other limits worth knowing
- Providers report their own coverage. The data is known to overstate availability. Where a provider tells you it cannot serve your address, it is right and the filing is wrong, which pushes the true monopoly figures higher still.
- Wireline means fiber, cable or DSL. Fixed wireless is excluded from the choice counts, because it competes on different terms and its coverage depends on signal rather than on plant. Satellite is excluded entirely: it reports essentially universal coverage, so including it would make every location in the country look competitive.
- Reporting service in a block is not an offer. Price, contract terms and installation feasibility are not in this data and we do not guess at them.
Who produced this
This analysis was written and computed by Ferenc Horvath, who runs WhereFiber as an independent project. It is not funded by, and carries no commercial relationship with, any internet provider or industry body, and no provider was given sight of it before publication.
The method is deliberately dull and reproducible. Every figure comes from the FCC's own Broadband Data Collection filing, period D25, published July 21, 2026, which anyone can download. We count, per census block, how many distinct companies report service, using the FCC's own provider identifiers rather than trading names, then aggregate. There is no model, no survey, no panel and no weighting. If you re-run the same counts on the same file you will get the same numbers, and if you get different ones we would like to hear about it.
Questions about the method, or a request for a cut we have not published: contact@wherefiber.com.
Reuse
These figures are derived from public federal data and you are welcome to cite them. A link back to this page is appreciated so readers can see the method. If you want a cut we have not published, by county, by metro or by operator, write to contact@wherefiber.com.
Related: fiber vs cable · fiber providers · cable providers · check a specific address