Cities Building the Most New Homes — Texas Edition — Study #3

Rainstone Water Heaters research program · Data: U.S. Census Bureau Building Permits Survey, final annual 2025 (released May 14, 2026) · Rates per 1,000 residents, computed against Census ACS 2024 5-year population · Analysis pulled October 22, 2026

The short version

We counted every new home authorized by building permits in Texas in 2025 — the Census Bureau’s Building Permits Survey, final annual figures — for every Texas place with 25,000 or more residents, and divided by population. Raw counts reward size; a rate per 1,000 residents shows where growth is actually fastest relative to the city that’s absorbing it.

Frisco leads its size class at 25.77 permits per 1,000 residents — roughly 26 new homes authorized in a single year for every 1,000 people who already live there. Among Texas’s biggest cities, Fort Worth builds fastest: 9,437 homes authorized in 2025, or 9.80 per 1,000 residents. And the pattern is unmistakable: the fastest building in Texas is happening in the suburbs ringing its big metros, not in the big cities themselves.

---

Key findings

• Frisco is #1 in the statewide 100,000–250,000 size class — 5,652 homes authorized in 2025, 25.77 per 1,000 residents. It also finishes #3 in the DFW metro and #4 statewide among all places of 25,000+, behind only three smaller boomtowns (Fulshear, Celina, and Princeton).
• Fort Worth is the fastest-building big city in Texas. At 9.80 per 1,000, it leads every place with 250,000+ residents — ahead of Austin (9.08) and Lubbock (8.99). Dallas (5.35) and Houston (5.10) build at barely half Fort Worth’s pace.
• Denton takes #2 in the size class at 22.49 per 1,000 (3,438 homes) — about 15% behind Frisco’s rate, and roughly 2.3 times the big-city leader’s pace.
• Georgetown is #2 in the Austin metro at 20.58 per 1,000 (1,770 homes), behind only Kyle (22.84).
• Schertz makes the San Antonio metro podium at #3 — 438 homes, 9.94 per 1,000.
• The scale gap is the story: the class leader’s rate (25.77) is about 2.6 times the big-city leader’s rate (9.80). Texas is growing fastest where its suburbs are.

---

The statewide field — what the data supports

Among all Texas places with 25,000 or more residents, the verified top of the statewide field is:

The rankings published in this study — every figure shown on this page, exactly as verified — are published with this study as a downloadable CSV.

An honesty note, printed on purpose: our verified dataset enumerates the top four of the statewide field by position, but carries the rate and unit count only for Frisco. Rather than publish figures we haven’t verified, the three smaller cities above Frisco are shown with their positions and no invented numbers — the statewide table in the downloadable CSV will carry their verified rates at page build, pulled directly from the BPS file. None of the three is a publisher in our Wave 1 census, so no outreach depends on their figures. The fully enumerated fields — the size classes and metro editions below — are where this study’s rankings live.

---

Size-class tables — where the podium finishes come from

Places with 100,000–250,000 residents

Ranks 5–9 in this class were not enumerated in the verified dataset and are shown as unranked in the full CSV rather than reconstructed here. Carrollton’s #10 is its verified class position.

Two North Texas neighbors take the top of the state’s most competitive class. Frisco’s margin over Denton is about 15% — decisive, not a rounding artifact — and both cities authorized homes at more than double the rate of any Texas city over 250,000 people.

Places with 250,000+ residents — the big cities

Dallas and Houston’s verified rates are published here; their exact positions below #3 were not enumerated in the verified dataset, so no rank is asserted for them.

Fort Worth’s 9,437 authorized homes is the largest absolute count in this study’s verified figures — and it pairs that volume with the big cities’ fastest rate. Austin’s #2 here is its second podium of the program (it also earned the most-improved recognition in our clean-water study) — a city building quickly while its water record improved.

---

Metro editions

Dallas–Fort Worth metro (verified positions)

Metro ranks #1–#2 were not enumerated in the verified dataset — the statewide top three (Fulshear, Celina, Princeton) sit in this metro’s orbit, but we do not assign them metro ranks our data didn’t state. Frisco and Denton’s metro positions are as verified.

Austin metro

Leander’s #3 position and Pflugerville’s #5 are verified; their rates were not carried in the verified dataset and are shown as “—” rather than estimated. Leander is a news-only publisher in our census (no honors page), so Georgetown is this metro’s podium publisher.

San Antonio metro

Seguin’s rate — nearly 24 per 1,000 — would contend in any field in the state. Neither Seguin nor Cibolo is a Wave 1 publisher, which makes Schertz, at #3, this metro’s podium publisher.

---

What rapid growth means for home services

Every one of these permitted homes is a household being built from the slab up — and every one of them gets a first water heater.

That’s the honest reason a water-heater company publishes this study:

• New construction is where tankless starts. A growing share of new Texas homes are built with tankless water heaters or plumbed to convert easily. The builder’s choice on day one becomes the homeowner’s replacement decision twelve to fifteen years later — growth suburbs are, quite literally, where our industry’s next decade of customers is being built right now.
• A new city’s service map is being drawn. The first plumber a new homeowner calls — for a warranty question, a first flush, a softener install — usually becomes their plumber. In a city adding 26 homes per 1,000 residents, thousands of those first-service relationships are forming every year.
• Growth cuts both ways for homeowners. Fast building can outrun the trades that support it. If you’re in one of these cities, the practical move is unglamorous: know where your water shutoff is, register your water heater’s warranty, and line up a local company before the first problem, not during it.

None of that changes the rankings above — they’re the Census Bureau’s numbers, and they stand on their own whether or not you ever need a water heater. If you’re in one of our Texas service areas and building, buying new, or maintaining what’s already installed, that day-to-day work is what Rainstone Water Heaters does.

---

Ranked city? Get your badge

If your city appears on a podium above, you’re welcome to post the result wherever your city posts honors like this — your news page, economic-development page, or awards page. We provide a ready badge (PNG and SVG, with a copy-paste embed), a one-page PDF summary of the study, the video at the top of this page as an embed, and a suggested social post. There’s no cost and no obligation, and your ranking stands whether or not you post anything. Contact the research team at [email protected].

Methodology in brief

Full methodology: published in full below. In short: U.S. Census Bureau Building Permits Survey, final annual 2025 (released May 14, 2026), place-level counts of new privately owned housing units authorized by building permits; rate = units authorized per 1,000 residents, using Census ACS 2024 5-year population for the same places as the denominator; Texas places with 25,000 or more residents; size classes and Census CBSA metro definitions as published. Two caveats we print rather than bury: a permit authorizes a home — it is not a completion certificate, and some authorized homes are built later or not at all; and this edition publishes only the ranks our verified data pull enumerated — positions it didn’t enumerate are marked, never reconstructed. One known data artifact: the BPS file shows Cedar Park with 0 units authorized in 2025, which is a reporting anomaly, not a real zero — Cedar Park is excluded from this study rather than ranked on a bad figure. If any figure for your city looks wrong, tell us — we publish corrections in a visible log.

FAQ

Permits aren’t the same as finished homes. Why rank on permits? Because the Building Permits Survey is the Census Bureau’s official, uniform, place-level measure of new-home building — every city counted the same way, in the same year. Permits are also the leading edge of growth: they show where building is happening now. We say “authorized” everywhere the distinction matters, and we never call a permit a completed home.

Why per 1,000 residents instead of total homes? Totals reward size — the biggest city would win every year by default. A rate shows intensity: how fast a city is growing relative to itself. We publish both: every podium figure on this page carries its absolute unit count alongside the rate wherever the verified data provides it.

Why the 25,000-resident floor? Below it, a single subdivision can double a tiny city’s rate, and the ranking becomes a lottery of small numbers. The floor keeps every ranked city a substantial place whose rate means something. The size classes (100,000–250,000; 250,000+) then group cities with genuine peers.

Why are some cells blank (“—”)? Because our verified pull enumerated those fields only partially — some positions without rates, some rates without positions. We publish what we verified and mark the rest. That is a deliberate choice: a study a communications office can check line by line is worth more than a complete-looking table with reconstructed numbers in it.

How often does this update? Annually, on the Census Bureau’s final BPS release. Rankings from different years are never mixed in one table, and prior editions stay archived.

Full methodology

This methodology was fixed in writing before the data pull was run (program spec, DATA-STUDY-PROGRAM.md §1, Study #4). The ranking rules below were not adjusted after seeing which cities won.

---

1. Data sources

• Permit counts: U.S. Census Bureau, Building Permits Survey (BPS) — final annual 2025 place-level file (released May 14, 2026). BPS publishes new privately owned housing units authorized by building permits at the permit-issuing place level (census.gov/construction/bps).
• Population denominators: U.S. Census Bureau, American Community Survey (ACS) 2024 5-year estimates, table B01003 (total population), for the same Census places.
• Analysis pull date: October 22, 2026. All figures in this edition come from that single analysis; figures from different data years are never mixed in one table.

2. Universe and population floor

• Included: Texas Census places with population ≥ 25,000 (ACS 2024 5-year).
• The floor exists so no rank rests on a tiny place’s volatile rate — below 25,000 residents, one subdivision permitting in a single year can double a rate, and the ranking stops measuring sustained building.
• Places below the floor are not ranked and are not contacted.

3. The metric

• Rate = (housing units authorized by building permits in calendar 2025 ÷ ACS 2024 population) × 1,000 — “new homes authorized per 1,000 residents.”
• Rates are published to two decimal places, exactly as computed in the analysis.
• Absolute unit counts are published alongside every rate wherever the verified dataset carries them. A rate without its count can mislead in both directions, so the pair travels together.
• The BPS count is of authorized units. A permit is not a start, a completion, or a certificate of occupancy; the study says “authorized” wherever the distinction matters and never describes a permitted unit as a built home.

4. Fields — classes and metros (fixed before the pull)

• Statewide field: all Texas places ≥ 25,000 population.
• Size classes (natural groupings, the same peer sets a communications office would use): places with 100,000–250,000 residents; places with 250,000+ residents.
• Metro fields use official Census CBSA definitions: Dallas–Fort Worth (19100), Austin–Round Rock–San Marcos (12420), San Antonio–New Braunfels (41700). Place-to-metro assignment is by place centroid within the CBSA.
• A city’s rank is asserted only in fields where the verified dataset enumerated that rank. This edition’s statewide field was enumerated only through position #4 (see §6).

5. Ranking rules

• Within every field, cities are ordered by permits per 1,000 residents, descending.
• No composite scores, no weighting, no adjustments. The metric is the ranking.
• Cities with genuine podium finishes in a field are the outreach set for that field (program rule: top 3 of every edition). Entities that are not verified publishers in the program’s publisher census appear in the published tables but receive no outreach this pass — the ranking is theirs regardless.

6. Partial-enumeration disclosure — read this before citing a rank

The verified analysis behind this edition enumerated some fields completely and others partially. Rather than reconstruct the gaps from secondary sources, this edition publishes exactly what was enumerated:

• Statewide field: positions 1–4 enumerated (Fulshear, Celina, Princeton, Frisco). Rates and unit counts verified only for Frisco (25.77 per 1,000; 5,652 units). The other three positions are published without rates or counts in this edition.
• 100,000–250,000 class: ranks 1–4 and rank 10 enumerated (Frisco 25.77 · Denton 22.49 · Conroe 18.49 · McKinney 14.43 · Carrollton 8.72). Ranks 5–9 not enumerated. Unit counts verified for Frisco (5,652) and Denton (3,438) only.
• 250,000+ class: ranks 1–3 enumerated (Fort Worth 9.80 · Austin 9.08 · Lubbock 8.99). Dallas (5.35) and Houston (5.10) rates verified; their positions below #3 not enumerated, so no rank is asserted for them. Unit count verified for Fort Worth (9,437) only.
• DFW metro: Frisco #3 and Denton #4 enumerated; metro ranks 1–2 not enumerated and not assigned to any city in this edition, even where statewide positions make a guess tempting.
• Austin metro: Kyle #1 (22.84), Georgetown #2 (20.58; 1,770 units), Leander #3, Pflugerville #5 enumerated; rates for Leander and Pflugerville not carried in the verified dataset.
• San Antonio metro: Seguin #1 (24.03), Cibolo #2 (12.19), Schertz #3 (9.94; 438 units) enumerated.

At page build, the complete BPS place file will populate a full downloadable CSV; any figure added there is pulled directly from that file and re-verified before publication — it will never be back-filled into this edition’s ranked tables without a dated correction note.

7. Known data artifacts and exclusions

• Cedar Park (0 units): the BPS place file shows Cedar Park with 0 units authorized in 2025. That is a reporting anomaly, not a real zero — a city of Cedar Park’s size did not literally authorize no homes. Cedar Park is excluded from this study entirely — neither ranked on the figure nor cited for it in either direction.
• Permit-issuing geography: BPS counts permits by the place that issues them. Where a city’s growth occurs in unincorporated county areas or under another jurisdiction’s permits, those units are counted by the issuing jurisdiction, not the city — a known feature of the dataset, disclosed here, that affects edge-of-city growth in particular.
• Non-reporting and imputation: BPS relies on permit offices reporting; Census imputes or carries forward for some non-responding offices in monthly files. This edition uses the final annual file, the most corrected vintage Census publishes.

8. Limitations — printed, not buried

• Authorized ≠ built. Some permitted homes are built in a later year; some are never built. The study measures building authorization, the Census Bureau’s uniform leading indicator — not completions.
• Single-year snapshot. One large master-planned subdivision’s permit batch can move a mid-size city’s annual rate. The rate is a 2025 fact, not a trend claim; the study makes no multi-year growth claims from a single year of data.
• Denominator timing. Rates divide 2025 permits by ACS 2024 5-year population estimates — the standard, published denominator for this metric, and itself an estimate with margins of error. In the fastest-growing cities the true 2025 population is higher than the 2024 ACS estimate, which if anything understates the denominator and overstates the rate slightly; that direction of bias is disclosed rather than hidden.
• This study measures construction volume only. It says nothing about affordability, home quality, infrastructure readiness, or whether growth is well-planned — and no ranked city’s finish should be read as a judgment on any of those.

9. Verification pass

Every rate, count, and rank published in this edition was cross-checked against the program’s verified analysis file (STUDY-MATCHES.md, computed from the BPS final 2025 place file and ACS 2024 on October 22, 2026). No figure appears in this edition that is not in that file; gaps in that file are published as gaps (§6), not filled.

10. Corrections policy

If any figure for your city looks wrong, contact the research team at [email protected]. Corrections are verified against the Census Bureau’s published BPS file, applied to the page and the CSV, and logged in a public corrections log with dates. A correction never requires a city to post, link, or pay anything.

11. Refresh and archiving

This study re-runs annually on the Census Bureau’s final BPS release. Prior editions remain archived at dated URLs. Rankings from different years are never mixed in one table.