How old are American homes?

The median American single-family home was built in 1979. It is 47 years old, it has outlived the shag carpet it was finished with, and there is a decent chance it still has the original ductwork. That figure comes from county assessor build years on 79,636,129 single-family homes — 93.6% of the 85.1 million in the national parcel record — derived August 30, 2026.

The mean is 50.8 years, higher than the median because the old tail is very long and the new one cannot be. A quarter of American homes were built before 1955; a quarter since 2001.

When America built its houses

Read this as a survival curve, not a construction record — it counts houses still standing, so early decades are undercounted by everything that has burned down or been demolished in the century since. The 1930s dip is real all the same: the Depression stopped American homebuilding almost completely, and 2,217,530 surviving homes from that decade sit below both the 1920s and the 1940s.

The peak is the 2000s at 11,099,247 homes, and what follows it is the more interesting number. The 2010s added 5,829,263 — a 47% collapse from the decade before, the housing crash written into the parcel record. The 2020s, at 4,166,289 homes through mid-2026, are running at roughly 625,000 a year against the 2000s’ 1.11 million. Construction has recovered from the crash and has not recovered from the boom.

Half the country predates the lead-paint ban

Lead-based paint was banned for residential use in the United States in 1978. 38,523,535 single-family homes — 48.4% of the dated stock — were built before that. That is the single most consequential number on this page, because it defines the population that federal renovation rules apply to and that anyone selling windows, siding, or a gut remodel has to work inside.

9,868,099 homes (12.4%) were built before 1940 — old enough for knob-and-tube wiring, plaster-and-lath, and no cavity insulation to be plausible defaults rather than surprises. 58,541,330 homes (73.5%) predate 2000 entirely.

Median home age by state

States with at least 200,000 single-family homes and at least 85% build-year coverage in the assessor record.

The spread is 44 years between the median New York home (1956) and the median Nevada home (2000). Nothing about the ordering is surprising, which is the point — a state ranking that reproduces the known shape of American settlement is a ranking you can trust for the cases where the answer is not already known.

Iowa at second place is the entry that repays attention. 32.6% of Iowa’s single-family homes predate 1940, the highest share of any state in this table and above New York’s 28.4%. Iowa built a great deal of housing before the war and comparatively little since; the stock did not turn over because the population did not grow. The same mechanism, in milder form, puts Nebraska and Kansas high on a list otherwise dominated by the industrial Northeast.

California at 1973 is younger than most people expect and only 9.4% pre-1940, despite the state’s reputation for expensive old housing. California’s scale is postwar: it added more homes in the 1950s through 1970s than most states have in total.

Seven states are excluded for assessor build-year coverage below 85% — Wisconsin (44.9%), Louisiana (48.3%), Maine (73.3%), South Dakota (74.0%), Arkansas (75.8%), Kentucky (76.0%) and Michigan (80.6%). They are named rather than silently dropped. Their build years are recorded for a large minority of parcels, but the counties that record and the counties that do not are not a random sample of each state, so a median computed on what remains would be a statement about assessor practice.

A third of the old stock is not as old as it looks

Assessors carry a second date alongside the build year: an effective year built, revised upward when a property is substantially rebuilt or renovated. On the 58,151,115 homes carrying both, 17,883,778 — 30.8% — have an effective year later than their actual one, with a median uplift of 16 years.

11,018,161 of those are pre-1978 homes that an assessor has judged materially newer than their build date. That is the renovation economy showing up in the tax record, and it cuts the practical read of the age numbers above: a 1955 house whose effective year is 1998 has had money spent on it, and it is a different prospect from its neighbour that has not. Neither is distinguishable by build year alone.

Methodology

• Source and cohort. County assessor records on Scout Data’s national parcel table, one row per property. 85,061,162 single-family homes, of which 79,636,129 (93.6%) carry a build year between 1800 and 2026. Condominiums, multi-family, mobile homes and land are excluded.
• Age is measured to 2026. A home built in 1979 is reported as 47 years old.
• Exact quantiles only. Medians and quartiles are computed exactly, not sampled. The sampled forms available in our query engine drift by up to half a year between identical runs, which is enough to reorder adjacent states; every quantile here was recomputed and matched before publishing.
• State coverage gate. A state appears only with at least 200,000 single-family homes and at least 85% build-year coverage. The seven excluded states are named on the page. Coverage does not drive the ranking — New York leads at 92.9% coverage while several states above 99% sit at the young end — which is the check that keeps this from being a table about assessor diligence.
• Survivorship. This counts homes standing today, not homes ever built. Pre-1940 figures are therefore lower bounds on what was constructed, and the further back a decade sits the more it understates.
• 2020s is a part decade — six and two-thirds years through August 2026 — and its share is not comparable to a full decade without annualising, which the text does where it matters.
• Effective year built is the assessor’s judgment about substantial renovation, and jurisdictions apply it inconsistently. The 30.8% figure is a share of homes where an assessor chose to move the date, not a measure of how many homes have been renovated.

Derived August 30, 2026. Spot something that looks wrong? Tell us — we’d rather correct a number than defend it.

Companion pages: roofing material statistics and roof replacement statistics cover what sits on top of this stock, HVAC replacement statistics covers what heats it, how long people stay in their homes covers who lives in it, and where people are moving covers where the new construction in the last row is going.

Frequently asked questions