Whole-house rewiring: when the wiring itself is the problem

Year built is a proxy for original construction — not a census of what is in the walls

The American Community Survey does not record conductor metal, jacket material, or whether a house has been rewired. Table B25034 records year the structure was built. That is a useful prompt: original construction in the United States used different wiring methods in different decades. It is not a count of remaining knob-and-tube, remaining cloth NM, or remaining aluminum branch circuits.

HyreElectrical calculation from the ACS 2024 1-year estimates, table B25034 (universe: housing units, United States), retrieved 5 September 2026: 146,740,960 housing units. Of those, 16,821,220 (11.5%) were built in 1939 or earlier; 6,210,969 (4.2%) in 1940–1949; 13,358,031 (9.1%) in 1950–1959; 14,113,943 (9.6%) in 1960–1969; and 20,140,924 (13.7%) in 1970–1979.

Those bands are the original-construction windows most often associated with the three eras on this page. Knob-and-tube was the dominant residential method into the 1930s and was still being installed, in some places, into the 1940s. Cloth-braid NM was listed in the NEC in 1926 and was ordinary construction through the 1950s and into the early 1960s. Solid aluminum branch wiring belongs to a much shorter window — CPSC’s pamphlet names 1965 through the mid-1970s.

None of those sentences is a remaining-stock count. Combined, units built in 1939 or earlier plus 1940–1959 total 36,390,220, or 24.8% of the ACS 2024 1-year stock — a pre-1960 original-construction proxy for ungrounded or early-grounded methods. Combined 1960–1979 units total 34,254,867, or 23.3% — a year-built proxy that includes millions of copper-wired houses and excludes any 1965–1973 aluminum that has already been remediated. Anyone selling “X million American homes still have knob-and-tube” or “still have aluminum” from ACS year-built is inflating a proxy into a census. We will not.

A 1912 house that was gutted in 1998 has modern NM in the walls and an ACS year-built of 1912. A 1971 house that has never been opened may still have original aluminum. The cable marking, the porcelain knobs, and a licensed electrician are the evidence. The year is the reason to look.

Year-built stock is the wrong denominator for remaining old wiring. Each bar is housing units still standing from those years, as of the ACS 2024 1-year estimates. Original construction method is a probability, not a finding.

Knob-and-tube: an ungrounded system that was designed to run in free air

Knob-and-tube is not “old Romex.” It is a different wiring method. Individual insulated conductors — a hot and a neutral, no equipment grounding conductor — are supported on porcelain knobs along framing, and they pass through framing in porcelain tubes. Splices in original work were made in free air, soldered and taped, not landed in a junction box. Where a conductor entered a box or a fixture, a woven cotton tube called loom covered the end. David A. Dini of Underwriters Laboratories, writing in 2006, records that the conductors were run spaced at least 2½ inches apart where exposed, and that knobs were not to support two wires of opposite polarity on the same insulator.

The physics of the method is the part later renovations break. Open wiring dissipates heat into free air. That is not a nostalgic flourish; it is why the model code still cares about insulation. NFPA 70, Article 394, still defines concealed knob-and-tube wiring and still permits it only for extensions of existing installations, or elsewhere by special permission from the authority having jurisdiction. Article 394.12 lists uses not permitted. Among them, in current editions: hollow spaces of walls, ceilings, and attics where those spaces are insulated by loose, rolled, or foamed-in-place insulating material that envelops the conductors. The 2017 NEC Handbook’s commentary on that prohibition is blunt: blown-in, foamed-in, or rolled insulation prevents dissipation of heat into the free-air space, resulting in higher conductor temperature, which could cause insulation breakdown and possible insulation ignition.

That is a model-code statement about new or extended concealed knob-and-tube in insulated cavities. Existing K&T that someone later buried in cellulose is a different, worse sentence: the original heat path is gone, and the rubber or cloth insulation on the conductors is a century old. CPSC staff put a number on the insulation-and-wiring combination in 1980. Release 80-002, 6 February 1980: staff estimated that more than 2,000 household fires occur every year when thermal insulation is ignited by faulty wiring or lighting fixtures, about three percent of the then-estimated 75,000 electrical-system fires annually. Staff’s mechanism, in the same release: the fire potential from damaged wiring or over-fused circuitry is greatly increased when the heat given off by the wiring — which normally is dissipated through free air circulation — is confined by thermal insulation. That 1980 estimate is not a 2026 remaining-K&T fire count. It is the federal safety agency stating the mechanism this page is using.

Dini records the same 1970s energy-crisis pattern from the laboratory side: loose-fill cellulose blown into attics and walls of older houses became a problem where knob-and-tube splices had glowing connections or overheated conductors. Glow is not a metaphor. Open splices can run hot for a long time without a box, without a cover, and without a ground-fault path.

The other structural fact is the missing equipment grounding conductor. There is no third wire. There is no armor to use as a ground path, as there is with Type AC. Three-prong appliances on a knob-and-tube circuit do not have an equipment ground, no matter what the receptacle looks like. Cheater plugs and three-prong replacements that are not GFCI-protected, not labeled, and not actually grounded are how that fact gets laundered into a listing photograph.

What the model code does not say is that existing knob-and-tube must be ripped out on sight. Article 394 exists because the method is still in buildings. New work in the method is limited to extensions of existing installations or special permission. Complete removal is an insurance, a renovation, a load, or a condition decision — not a sentence the NEC prints as a national mandate. InterNACHI and a great deal of contractor copy collapse that distinction. This page will not.

The knob-and-tube insurance conversation is not the aluminum conversation

There is no national underwriting rule we can cite. Carrier manuals are not a federal dataset. The honest move is the same one the aluminum child page makes: quote a named carrier, date the retrieval, and tell you to get your own carrier’s answer in writing.

Citizens Property Insurance Corporation is Florida’s residual-market carrier. Its public “Home Condition Requirements” sheet for new-business HO-3, HO-8, DP-1 and DP-3 (dated 1 March 2023, retrieved 5 September 2026) lists knob and tube wiring under Unacceptable Conditions, alongside “fewer than 100 amps,” “aluminum branch circuit wiring,” and a list of hazardous conditions (inadequate fusing, open circuits or missing knockouts, frayed or loose wiring, double taps). Acceptable conditions on the same sheet include 100 amps or more, circuit breakers/220 volts, fuses, and multi-strand aluminum wiring.

Read the asterisk. Aluminum branch circuit wiring carries a footnote: upload documentation that the property was remediated using a method acceptable to Citizens, and see Uninsurable Properties in the underwriting manual. Knob and tube wiring does not carry that footnote on this sheet. Aluminum has a published remediation path at Citizens — COPALUM, AlumiConn, or a full copper rewire, as the carrier’s separate FAQ (Answer ID 1550, updated 16 April 2026) spells out. Knob-and-tube, on the condition sheet, is listed as unacceptable, full stop.

HyreElectrical analysis: that is the information-gain of scoping by era. A household that treats “old wiring” as one underwriting problem will hear “CPSC-approved connectors” in an aluminum conversation and assume the same paper works for knob-and-tube. It does not, at this named carrier, on this dated sheet. COPALUM is an aluminum-to-copper connection repair. It has nothing to do with porcelain knobs. Asking for a “rewire or connectors, whichever is cheaper” on a K&T house is asking the wrong menu.

Citizens is one Florida residual-market carrier. Florida is relevant on this site because 74% of our contractor records are there, which we have already disclosed, and because Citizens publishes the rule. It is not a national fact that “insurers refuse knob-and-tube.” Some admitted carriers in other states will write a house with documented, inspected, uninsulated K&T; some will not write it at any price. Get the answer from the carrier that will actually bind the policy, named method, in writing, before you pay for one.

A 4-point inspection in Florida is how this conversation usually starts. Citizens requires a completed 4-point form for homes over 20 years old, verifying that the electrical system is in an acceptable condition and has sufficient amperage, and says: if an inspector indicates the electrical system is not in good working order, do not submit without proof of repairs or updates. Active knob-and-tube is not an “acceptable condition” on the sheet above. That is why a sale, a refinance, or a change of carrier is a common trigger for a K&T rewire even when the lights still work.

Cloth-insulated cable is the generation after knob-and-tube — and it is not one product

People say “cloth wiring” as if it were a single SKU. It is at least two methods, and the grounding question splits each of them in half. Getting this wrong is how a house is either over-rewired or left ungrounded with a fresh coat of three-prong receptacles.

Cloth-braid NM is early nonmetallic-sheathed cable: two (later three) rubber-insulated conductors inside a woven cotton or rayon jacket, often tarred or varnished. Dini: NM was first listed and described in the NEC in 1926; General Cable’s Rome Wire Division had invented it a few years earlier as Romex. Until the early 1960s, most NM for residential use did not have a grounding conductor. Around 1950, synthetic spun rayon was permitted to replace cotton in the braid. In the early 1960s thermoplastic began replacing the braid; by about 1970 most NM had a PVC jacket, though a braid was still permitted until 1984. That means a house built in 1958 and added-onto in 1972 can have cloth-braid ungrounded NM in the original rooms and plastic-jacketed NM with a ground in the addition. ACS year-built will report 1958.

Armored cable — BX, Type AC is the other cloth-era method people photograph and misname. Sprague Electric listed flexible steel-armored conductors in 1899. The production version was “BX”; the name stuck as a GE trade name. AC first appeared in the 1903 NEC and, Dini says, became popular around 1930, the same period knob-and-tube was waning. The armor can serve as an equipment grounding path; AC made after 1959 requires an aluminum bonding strip under the armor to improve that path. Older AC without the bonding strip is a different grounding conversation from cloth NM with no ground at all. The Fire Protection Research Foundation’s 2008 aging study found older armored cable without the bonding strip with armor resistances that in some cases were more than double the original design value. That is a grounding-path finding, not a “cloth is cloth” finding.

The insulation on the conductors, in both methods, was gum-rubber for a long time — a sulfur-cured compound that required the copper to be tinned because the sulfur attacked bare copper. Dini: during the 1950s the industry began transitioning residential wire insulation from rubber to thermoplastics; PVC does not suffer from the brittleness and cracking with age that was typical of the older rubber. The FPRF aging project, harvesting wiring from 30 houses, made the same split: thermoset rubber from before the 1950s was still performing in many residential environments, but older rubber compounds become brittle and are a potential hazard when subjected to bending, abrasion, or harsh usage; thermoplastic insulated wires typical of the 1950s and later continued to perform with excellent results under most conditions. Condition of the rubber, not the word “cloth” on a 4-point form, is the finding.

Then the 1962 line. Dini, from the NEC: in 1962 the Code was revised to require all branch circuits to include a grounding conductor or ground path to which the grounding contacts of the receptacle must be connected. That effectively discontinued non-grounding-type receptacles except for replacement use in existing installations where a grounding means might not exist. Earlier versions of NM-with-ground permitted an undersized grounding conductor (No. 16 AWG ground on 14 and 12 copper NM); 1969 ended that for 14, 12 and 10 gauge NM. So “early grounded cloth” is a real product: cloth-jacket or early plastic NM that does contain a ground, sometimes an undersized one. Treating every cloth jacket as ungrounded is how people miss a ground that is already there — and treating every cloth jacket as fine because “it has a ground” is how people miss brittle rubber on the current-carrying conductors.

The cloth insurance conversation is about condition — and it is not knob-and-tube

Citizens’ 1 March 2023 Home Condition Requirements sheet names knob-and-tube and aluminum branch wiring as unacceptable conditions. It does not name cloth-jacket rubber-insulated cable as a third uninsurable category on that sheet. Fuses are listed as acceptable. Hazardous conditions — inadequate fusing, frayed or loose wiring, double taps — are listed as unacceptable. A 4-point inspection that reports damaged cloth insulation, overfusing, or other hazardous findings is a different paper from a 4-point that reports intact cloth NM with no hazardous conditions.

That is not a prediction that every carrier will write a cloth-wired house. It is a description of one dated, named sheet. Other Florida carriers, and carriers in the other four states in our store, may treat cloth as a decline, as a surcharge, or as a documentation problem. We did not find a second named, dated, public underwriting FAQ of Citizens-level specificity that defines cloth the way Answer ID 1550 defines aluminum, and we will not fill that gap from broker blogs. Ask the carrier that will bind the policy what the 4-point must show.

HyreElectrical analysis: the practical error is using “cloth” as a synonym for knob-and-tube in an insurance email. Knob-and-tube is a named unacceptable condition at Citizens. Cloth is a description of a jacket. An inspector who writes “cloth/K&T” as one checkbox is collapsing the three-era table on this page. Ask for the method to be named.

Aluminum branch wiring is a connection problem — and a full copper rewire is only one of three CPSC methods

Aluminum branch wiring has its own child page, and this parent will not duplicate it. The depth on CPSC Publication 516, on COPALUM versus AlumiConn versus copper replacement, and on what Citizens will accept as documentation, lives at aluminum branch wiring remediation. What belongs here is the parent-level distinction: aluminum is the era in which a recognized repair is not replacement.

CPSC Publication 516 (June 2011, still the current numbered pamphlet as of our 5 September 2026 retrieval) names three methods it will treat as a permanent repair of residential aluminum branch-circuit wiring: complete replacement with copper cable; COPALUM crimp pigtailing at every connection; and AlumiConn set-screw pigtailing, which staff call the next-best alternative if COPALUM is not available locally. Twist-on connectors and CO/ALR devices are not, in CPSC’s published view, a permanent repair. COPALUM and AlumiConn leave the original aluminum in the walls. Only a copper rewire removes it.

That is why a sales pitch that jumps from “the house has aluminum” to “you have to rewire” is skipping CPSC’s own ranking. It may still be the right job — walls already open, a carrier that will not accept connectors, a remodel that was going to happen anyway. It is not the only CPSC-recognized job. Citizens, again, will accept documented COPALUM or documented AlumiConn as alternatives to a full copper rewire. Other carriers may not. The order of operations that avoids paying twice is on the child page: insurer and lender in writing, then the electrician quotes that named method.

CPSC’s identification advice is the homeowner action: read the markings on the cable jacket in unfinished basements, attics or garages. Aluminum conductors are marked “Al” or “Aluminum” every few feet. “CU-clad” or “Copper-clad” is a different product and is excluded from those repair recommendations. Do not open the panel to check. Year built (roughly 1965 to the mid-1970s) is a reason to look, not proof. The ACS 1960–1979 band on this page is a year-built proxy, and it is an order of magnitude larger than CPSC’s two-million estimate of homes built with the product. The child page does that arithmetic with the 2019–2023 5-year vintage; the bands here use the 2024 1-year vintage. Both are proxies. Neither is remaining aluminum.

A new panel does not remediate aluminum connections in the rooms. A Federal Pacific or Zinsco load center is a different era-hazard and can coexist with aluminum branch wiring; that identification lives on replacing a Federal Pacific panel.

A panel upgrade is a different job. So is adding a charger.

The load center is the bus, the breakers, the enclosure, and often the service rating. The branch wiring is the cables in the walls. Replacing one does not remediate the other. A 200-amp panel on knob-and-tube is a new lid on an old method. An aluminum-wired house with a new Square D load center still has aluminum at the receptacles. The decision rule for the panel — calculated load, hazard-era product, damage, or a utility service change — lives on the panel-upgrade page. Cost numbers for that job live on panel-upgrade cost. This page will not import them and will not invent a combined “rewire plus panel” national figure.

Spaces and amps fail independently there, too. A 200-amp panel can be full; a 100-amp panel can have empty slots and still fail a load calculation. Count spaces on the panel space checker. Run a neighborhood load check on the electrical load calculator, which is labeled as not a stamped calculation. If the reason you are in the walls is a heat pump and a charger, the load conversation and the wiring-era conversation have to happen in the same week, not as two surprises.

Federal Pacific Stab-Lok and Zinsco / Sylvania-Zinsco are named panel products with a published record. They are not a synonym for old wiring. Identification is its own page. A house can have a Stab-Lok and cloth NM, or a later panel and aluminum branch, or a Stab-Lok and knob-and-tube. Three findings, possibly three invoices.

What actually changes the price — and why this page will not ship a national average

Contractor blogs publish round numbers for “rewire a house.” Those round numbers are not a dataset. They do not say which era, whether the walls are open, whether the plaster is historic, whether the attic is finished, whether the panel is in the scope, whether the utility is involved, which edition of the code the city has adopted, or who patches the drywall. Labour rates in Miami-Dade and labour rates in Salt Lake City are not one market. We do not have a sourced, dated survey of completed whole-house rewire invoices, and we will not borrow a blog range to fill the hole. Type the bid. Compare bids only when they share a scope.

What moves a real bid, in terms an electrician can put on a page:

The era, because the method is the labour. Pulling new NM through an unfinished basement and abandoning dead K&T is a different number of hours from fishing every run in a plaster-and-lath two-story with a finished attic. COPALUM at every device, if that is the aluminum job, is a connection-count job, not a pull-count job. Treating those as one “per square foot” is how a cheap bid and an honest bid cannot be compared.

Access. Unfinished basement and accessible attic versus plaster ceilings, lath, brick, and no chase. Occupied house versus vacant. Who repairs the finish — the electrician, a drywall contractor, you — has to be in the scope, or it will show up as a second invoice the week the inspector passes the electrical.

How much of the house is live on the old method. A 1910 house that was already rewired except for two attic circuits is not a whole-house rewire. A 1968 house with aluminum in every wall is. ACS year-built will not tell the electrician which one you are. Opening unfinished space will.

Whether the panel, the service, or both are in the same mobilization. If the calculated load also fails, or the load center is a hazard-era product, doing the wall work and leaving the Stab-Lok is leaving a named problem in place. Combining the jobs can be cheaper than two mobilizations. It can also be two permits and a utility appointment. See the panel-upgrade page for the second bill the utility sometimes sends. This page will not invent a combined price.

Permit and inspection. Named local examples are below. They are a line on the invoice, not the invoice.

Documentation the insurer or lender has asked for. A letter that no knob-and-tube remains in service, photographs of abandoned ends, a named COPALUM or AlumiConn packet, a passing inspection. An invoice that says “pigtail outlets” without naming the method is how a 4-point fails six months later. That paperwork is labour. It should be in the scope.

If a bid is a single number with the word “rewire” on it, it is not yet a bid you can compare. The FTC’s guidance on hiring a contractor is to get itemised estimates in writing; allowances are placeholders. The same rule applies here.

Weatherization, historic houses, and why “just insulate around it” is a local-code question

An energy retrofit is one of the most common ways live knob-and-tube gets buried. It is also a place where the model code and a handful of state amendments disagree, and where a federal weatherization program has spent forty years telling states to follow the local electrical code rather than a DOE national rule.

In 1987 the NEC (then Section 324-4) was revised to prohibit concealed knob-and-tube in hollow spaces insulated by loose or rolled material; later editions add foamed-in-place. DOE’s Weatherization Assistance Program restated that change in program guidance and, in an 21 October 1988 memorandum that still appears in later archived WPNs, put responsibility on the states to ensure that insulation around knob-and-tube conforms to the applicable codes in the jurisdiction where the work is being performed. Washington, Oregon, and some local jurisdictions have amended the model rule. Washington Administrative Code 296-46B-394, for example, does not prohibit loose or rolled thermal insulation in spaces containing existing knob-and-tube if a licensed electrical contractor surveys the wiring and certifies it is in good condition, the insulation meets Class I specifications with a flame spread of 25 or less, foam is not used, and overcurrent protection complies with the 60°C column of the ampacity table via circuit breakers or Type S fuses.

That is a named state amendment, not permission to blow cellulose over live K&T in Miami-Dade. Florida did not adopt Washington’s exception by virtue of this paragraph. If you are insulating, the sequence is: licensed electrician identifies whether the K&T is live; the adopted electrical code, including any local amendment, says whether insulation may envelop it; then the insulation contractor works. Reversing that sequence is how a weatherization job becomes an electrical fire conversation. CPSC staff said as much in 1980, before the 1987 code change: have the wiring inspected in areas where insulation is to be installed; look first for frequent fuse-blowing, flickering lights, or hot outlet covers.

Historic fabric is the other constraint the square-foot blogs skip. Fishing NM through plaster-and-lath without destroying the finish is slow. Surface raceway, or opening one face of a wall on a room-by-room remodel, are different jobs from a gut. A preservation commission may have a view on visible work; the building department still has a view on concealed work. Neither is a reason to leave damaged live conductors in place. Both are reasons the bid has to describe access.

Where we actually have records

HyreElectrical is an information site and a matching project. It is not an electrical contractor, and it does not have a service radius. What we do have is a contractor store we counted on 5 September 2026: 16,369 companies across 5 states — Florida 12,114, Utah 1,489, Nevada 1,066, California 903, Texas 839. 74% of those companies are in Florida. The other 45 states are not in this store. Nothing here is a national directory of electricians, and a page that used these counts as one would be misusing them.

If you are in those five states, the useful next step is a licensed electrician who already pulls permits in your jurisdiction, checked against the board that issued the license. If you are not, the same hiring questions apply and our store has nothing extra to add. The enquiry form is how matching, as it is built, will know which records are relevant.

HyreElectrical contractor store, counted 5 September 2026 from company-state records. A research sample in five states, not a census of the trade.

What the fire numbers actually measure — and what they do not

Old wiring is a fire conversation, and it is a conversation in which round numbers get attributed to the wrong product. This page will use the federal and NFPA figures as they are published, with their denominators.

CPSC’s 2019–2021 Residential Fire Loss Estimates (July 2024 staff report, retrieved 5 September 2026) put electrical-distribution equipment at an estimated annual average of 17,300 residential structure fires, 140 deaths, and 640 injuries over those three years. Installed wiring, a subset, averaged 7,600 fires, 50 deaths, and 230 injuries a year. Receptacles and switches averaged 2,300 fires; panel boards 700. Those are equipment-involved estimates from NFIRS and NFPA’s fire-department survey. They are not a count of knob-and-tube fires, cloth-NM fires, or aluminum-connection fires. CPSC does not publish that split.

NFPA Research, in a 1 May 2026 report by Tucker McGree, estimated that in 2020–2024 local fire departments responded to an annual average of 46,652 home electrical structure fires — a broader net that includes fires where electrical failure or malfunction was a factor contributing to ignition, or where electrical distribution, lighting, or power-transfer equipment was involved. Those fires caused an estimated annual average of 527 civilian deaths, 1,580 civilian injuries, and $2.4 billion in direct property damage, and accounted for about 13% of home structure fires and 18% of home structure-fire deaths. That is a different definition from CPSC’s electrical-distribution equipment table. Both can be true. They are not interchangeable.

Age of housing is a risk factor in the federal record, with a sample caveat that copy usually drops. A 10 March 2003 CPSC economics memorandum on AFCIs, citing the 1990 CPSC epidemiological study Residential Electrical Distribution System Fires, states that 85% of all such fires involved housing over 20 years old — and, in the footnote, that the study was based on 149 investigated fires in 16 cities and does not represent a statistically representative sample. CPSC’s 19 May 1993 news release, selecting home electrical-system fires as a 1995 priority, said data showed a disproportionately high frequency of electrical-system fires in homes more than 40 years old, and that about one-third of US homes were then more than 40 years old. Those are CPSC’s sentences. They are not a relative-risk ratio we can compute from ACS year-built, and they are not a claim that knob-and-tube causes a stated share of the 7,600 installed-wiring fires.

The useful conclusion is narrower. Installed wiring is a real, counted slice of residential fire loss. Older housing is over-represented in the electrical-distribution fire investigations CPSC has published, on a sample that CPSC itself flags as not nationally representative. The mechanism for knob-and-tube-plus-insulation is independently stated by CPSC staff (1980) and by the NEC handbook commentary on Article 394.12. The mechanism for aluminum is independently stated in Publication 516. The mechanism for aged rubber is independently stated by UL’s Dini and by the FPRF aging project. None of those mechanisms needs a fabricated “K&T causes 30,000 fires a year” statistic, and we have not found a CPSC or NFPA table that supports one.