SF-0017Strange America
Lead began Picher's relocation. Collapse risk widened it. A tornado hastened it
Oklahoma's 2004 relocation covered only families with children under seven. Then the Army Corps found a third or more of homes had collapse potential. Then, in 2008, an EF4 tornado hit the town.
A chat pile on the St. Joe property at Picher — crushed rock left from lead and zinc milling, piled above ground. The ruined concrete structure at the left gives the scale.
Panoramic view of a chat pile on the St. Joe property, Picher, Oklahoma, looking east.
- Image
- S. Jerrod Smith, US Geological Survey
- Taken
- 2013
- Location
- Picher, Ottawa County, Oklahoma
- Archive ref.
- USGS media library — Oklahoma-Texas Water Science Center
- Licence
- Public domain per the USGS media library's item-level Sources/Usage statement (licence record)
- The image itself
- Authentic, unaltered
Panoramic view of a chat pile on the St. Joe property, Picher, Oklahoma. S. Jerrod Smith, US Geological Survey, Oklahoma-Texas Water Science Center, 2013. Public domain. Cropped.
Verdict
"America's most toxic town — abandoned because the lead poisoned the children."
The core of the claim holds up, but a specific detail as it circulates is wrong or overstated.
The claim names one cause where the record describes a sequence, which is what this verdict is for.
TRUE, AND FIRST: lead and other metals are why Tar Creek exists as a Superfund site. EPA proposed the site for the National Priorities List on 27 July 1981 and added it on 8 September 1983 after elevated cadmium, lead and zinc were found. Blood-lead testing in August and September 1996 found 38.3 percent of children tested in Picher — 31 of 81 — above 10 micrograms per decilitre. Lead drove the first relocation: in May 2004 Oklahoma implemented a relocation program for families with children under the age of seven. Lead began Picher's relocation.
SECOND, AND WIDER: a US Army Corps of Engineers study, requested by Senator James Inhofe, "found that a third or more of the homes in the area have varying degrees of collapse potential due to extensive historic mining under the towns of Picher, Cardin, and Hockerville". A public trust established by the State of Oklahoma then ran a voluntary relocation specifically "for properties in Cardin, Picher, and Hockerville that are impacted by potential subsidence", under conditions set by the legislature in May 2006. Collapse risk widened it.
THIRD, AND FASTER: on 10 May 2008 an EF4 tornado with estimated winds of 165 to 175 mph struck Picher, killing six people there. Eleven days later EPA announced $8 million to Oklahoma's Department of Environmental Quality under the headline "EPA Provides Funds To Expedite Buyout of Picher Residents", to "expedite the buyout and relocation of residents" in what it called the disaster area. The word is the agency's own. A tornado hastened it.
What this verdict does not do is rank the three, or add a fourth from memory. Picher's population history before 2004 is not established by anything read here and no trajectory is asserted.
Mostly True requires 1 primary or authoritative source. This file cites 5.
In short
Picher, Oklahoma sat on the Tar Creek Superfund site, listed in 1983 for acid mine water and heavy metals. No single cause emptied it. Lead drove a narrow 2004 relocation of families with children under seven. A US Army Corps of Engineers study then found a third or more of the homes had varying degrees of collapse potential from the workings beneath them, and a state trust bought out properties impacted by potential subsidence. On 10 May 2008 an EF4 tornado struck the town, and eleven days later EPA released funds "to expedite the buyout". Picher and Cardin were dissolved.
What the record actually says
Picher sat on the Oklahoma end of the Tri-State Mining District, where lead and zinc were mined from the early 1900s until mining “ceased in the region by 1970”. The milling left chat — crushed rock — in piles above ground, and it left several hundred feet of worked-out rooms and pillars underneath the town.
The environmental case came first. When the pumps stopped, the mines refilled, and in 1979 acid mine water began discharging into Tar Creek. EPA proposed the site for the National Priorities List on 27 July 1981 and added it on 8 September 1983.
Then the children. Blood-lead testing in August and September 1996 found 38.3 percent of the children tested in Picher — 31 of 81 — above 10 micrograms per decilitre. In Cardin it was 62.5 percent, though of only 16 children.
That is the story everyone knows, and all of it is true.
Lead began it
In May 2004, Oklahoma set up a relocation programme. It covered families with children under the age of seven. That is a public-health intervention, and it is not a town closing — but it is where the moving starts.
Collapse risk widened it
What changed the scale was a study. At the request of Senator James Inhofe, the US Army Corps of Engineers looked at the ground, and EPA summarises the result in a single sentence:
a study that found that a third or more of the homes in the area have varying degrees of collapse potential due to extensive historic mining under the towns of Picher, Cardin, and Hockerville
A public trust established by the State of Oklahoma then ran a voluntary relocation — and EPA is specific about what it was written against — “for properties in Cardin, Picher, and Hockerville that are impacted by potential subsidence”, on conditions set by the legislature in May 2006. The CERCLA buyout followed with the OU4 Record of Decision in February 2008.
A tornado hastened it
At about 5:38 in the afternoon of 10 May 2008, an EF4 tornado moved into Picher. The National Weather Service puts its winds at “165 to 175 mph”, its northeast Oklahoma track at “approximately 29 miles, with a maximum width of one mile”, and its toll in the town at six people killed.
Eleven days later, EPA published a news release. The headline is the finding:
EPA Provides Funds To Expedite Buyout of Picher Residents
Eight million dollars went to Oklahoma’s Department of Environmental Quality — three million directed by Congress for relocating residents of Picher, Cardin and Hockerville, five million through Superfund — to “expedite the buyout and relocation of residents” and help with “demolition or relocation of homes, businesses, and public use structures located in the disaster area.”
Expedite, not begin. The buyout the tornado accelerated had been running for two years.
Demolition of the purchased homes began on 24 January 2011. Oklahoma’s Department of Environmental Quality issued a closeout report on 23 November 2011. Picher and Cardin, EPA records, “were dissolved”.
Three things, not one
It would be tidy to end with so it wasn’t the lead after all, and that would be a worse sentence than the one it replaces. It would be equally tidy — and this file made the mistake in its first version — to say the ground did it instead.
Lead is why there was a Superfund site to begin with. Lead is why EPA remediated roughly 2,887 residential yards and public properties. Lead is why elevated blood-lead levels among screened children at Tar Creek fell from 21.5 percent in 1997 to zero by 2013 — a genuine public-health result, achieved while the town around it was being bought and knocked down. And the CERCLA buyout was implemented by a body called the Lead-Impacted Communities Relocation Assistance Trust.
Even the paperwork points at lead. The subsidence finding is what the general buyout was written against. The tornado is what the federal money was released to speed up.
None of the three is the answer. The sequence is the answer.
What this file does not claim
That lead was incidental at Picher — it was not. That collapse risk alone emptied the town; an earlier version of this file said so in its headline, and that was the same single-cause error with a different culprit. That the tornado started the evacuation: the dates rule it out and EPA’s own verb is “expedite”.
That the two relocation programmes were the same programme, or run by the same trust; EPA names LICRAT for one and not the other, so neither does this file. That any particular number of people left for any particular reason: the buyout totals are property counts, not motives. That the Corps study said more than EPA quotes it as saying — the study itself was not retrieved.
And not that Picher had been emptying for decades before any of this. It is a reasonable thing to assume about a mining town whose mines closed by 1970, and no source read here gives a population figure for Picher in any year, so this file does not put one in the reader’s head.
The evidence
Tar Creek was listed for metals, not for subsidence. EPA proposed the site for the National Priorities List on 27 July 1981, and "the Site was added to the NPL on September 8, 1983" after a remedial investigation "showed elevated concentrations of metals (cadmium, lead and zinc) in the mined areas of Ottawa County."
The lead exposure was real and measured. "Blood lead tests conducted in August 1996 and September 1996 found that 38.3 percent (31 of 81) of the children tested in the community of Picher had blood lead concentrations exceeding 10 micrograms per deciliter (µg/dL), that 62.5 percent (10 of 16) of the children tested in Cardin had blood lead concentrations exceeding 10 µg/dL, and that 13.4 percent (9 of 67) of the children tested in Quapaw" exceeded it.
Cited sources:S2
The lead-driven relocation was narrow. "In May 2004, the State of Oklahoma implemented a relocation program for families with children under the age of seven." It did not cover the town.
Cited sources:S2
The finding that changed the scale was structural. "At the request of U.S. Senator James Inhofe, the U.S. Army Corps of Engineers conducted a study that found that a third or more of the homes in the area have varying degrees of collapse potential due to extensive historic mining under the towns of Picher, Cardin, and Hockerville."
Cited sources:S2
The general buyout was written against subsidence, not lead. "A public Trust established by the State of Oklahoma implemented a voluntary relocation for properties in Cardin, Picher, and Hockerville that are impacted by potential subsidence. The Trust offered residents to voluntarily apply for relocation under conditions established by the State Legislature in May 2006."
Cited sources:S2
Subsidence was already a named federal concern three years before the Corps study. A Memorandum of Understanding signed in May 2003 between EPA, the Department of the Interior and the Army Corps "provides for a coordinated response to environmental contamination, physical safety concerns (open mine shafts, subsidence, and flooding), and poor economic conditions."
Cited sources:S2
The towns did not merely empty; they were dissolved. EPA records that Picher and Cardin "were part of the site, but were dissolved as part of a residential relocation and buyout effort", and that "Picher, Cardin and Hockerville were part of the voluntary buyout of properties and are currently uninhabited."
Cited sources:S2
Demolition and closeout are dated. "The demolition of the homes purchased in the towns of Picher, Cardin, and the Hockerville began on January 24, 2011, and has been completed. ODEQ described this work in a Closeout Report issued November 23, 2011."
Cited sources:S2
A second, later buyout ran under an explicitly lead-named body. The OU4 Record of Decision signed in February 2008 included "relocation, which will continue to be implemented by the Lead Impacted Communities Relocation Assistance Trust (LICRAT)", extended in February 2010 to include Treece, Kansas. EPA's site profile records the buyout as covering "628 residences, 74 businesses, and 125 renters."
The mining that undercut the towns had stopped decades before. "Extensive lead and zinc mining began in the early 1900s. Mining ceased in the region by 1970." The acid mine water that first drew attention appeared in 1979, when pumping had stopped and the workings refilled.
Cited sources:S2
A tornado struck the town while the buyout was already running. The National Weather Service rates it "EF4 on the Enhanced Fujita Scale with winds estimated at 165 to 175 mph", records that it "resulted in 6 fatalities", and gives a track in northeast Oklahoma of "approximately 29 miles, with a maximum width of one mile", with "extensive and widespread damage" to residential structures. Radar imagery on the NWS page is timestamped 5:38pm as the tornado moves into Picher.
Cited sources:S4
EPA linked the tornado directly to the pace of the buyout, in its own headline: "EPA Provides Funds To Expedite Buyout of Picher Residents", dated 21 May 2008. The release states the town "was heavily damaged by a May 10 tornado", and allocates $8 million to the Oklahoma Department of Environment Quality — $3 million directed by Congress for relocation of Picher, Cardin and Hockerville residents and $5 million through Superfund — to "expedite the buyout and relocation of residents" and assist with "demolition or relocation of homes, businesses, and public use structures located in the disaster area."
Cited sources:S5
EPA's post-tornado sampling did not find a contamination emergency. The agency deployed emergency response staff to "conduct air monitoring and soil sampling", and reported analysis showing "no health risks to residents or responders from the debris created by the tornado."
Cited sources:S5
The lead problem was substantially remediated while the town was being removed. EPA records elevated blood lead in screened Tar Creek children falling "from 21.5% in 1997 to 0% in 2013", and an ATSDR finding that children aged 1 to 5 above 10 µg/dL "decreased from 31.2% in l996 to 2.8% in 2003."
Cited sources:S2
Evidence that cuts the other way
Findings that complicate or argue against the verdict above. They are listed because leaving them out would make the file look stronger than it is.
Lead did not stop being the reason for the federal presence. The site is a Superfund site because of metals; the OU2 residential soil programme remediated roughly 2,887 residential yards and public properties; and the trust that ran the later buyout is named for lead-impacted communities. A reader who came away thinking lead was incidental at Picher would be more wrong than one who thought it was the whole story.
Cited sources:S2
No single cause emptied Picher, and an earlier draft of this file said one did. Its headline read "Mine collapse is what emptied it", which is the same error as the claim being corrected with a different culprit substituted. The record supports a sequence — a lead-driven programme for families with young children, a wider subsidence-driven buyout, and a tornado after which EPA funded expediting what was already under way — and no source read here weighs them against each other. The correction is recorded rather than quietly edited.
The tornado did not begin the evacuation, and the dates settle it. The Corps study and the state trust's subsidence relocation ran from conditions set in May 2006, and the OU4 Record of Decision was signed in February 2008 — both before 10 May 2008. EPA's own word for what the post-tornado money bought is "expedite", not "initiate".
The two relocations should not be collapsed into one. The subsidence relocation ran under a public trust on conditions set by the Oklahoma legislature in May 2006 and closed out in November 2011; the OU4 relocation was a CERCLA remedy component signed in February 2008 and implemented by LICRAT. The EPA document names LICRAT for the second and does not name the trust in the first passage, so this file does not assert that they are the same body.
Cited sources:S2
EPA's own blood-lead figures differ between documents and years, and the file quotes both rather than the most dramatic. The site profile reports that in 1994 Indian Health Service testing found "approximately 35 percent of the Indian children tested" above 10 µg/dL; the 2014 community involvement plan reports 38.3 percent of children tested in Picher in August–September 1996. These are different populations, different years and different testing programmes.
What we still don’t know
- The date and full findings of the Army Corps subsidence study. EPA's community involvement plan reports the study's headline conclusion and that Senator Inhofe requested it, but does not date it, name it, or give its methodology. The study itself was not retrieved, so nothing beyond the quoted conclusion is asserted — in particular, "a third or more of the homes" is EPA's summary of the Corps' finding rather than a figure read from the Corps' own report.
- Whether the trust that ran the subsidence relocation and LICRAT are the same body. EPA names LICRAT for the OU4 relocation and refers to "a public Trust established by the State of Oklahoma" for the subsidence relocation without naming it. They may well be the same trust; the document does not say so and this file does not.
- How many people left for which reason. The buyout totals are recorded as property and household counts across programmes, towns and states, and no source consulted breaks participation down by whether a household qualified on subsidence grounds, on lead grounds, or both. The file therefore describes what each programme was written against, not how many people each moved.
- Picher's population history, which this file does not establish and deliberately does not assert. It is often said that the town had been shrinking for decades as mining wound down, and that is plausible — EPA records that "mining ceased in the region by 1970" — but no source read here gives a population figure for Picher in any year. A trajectory that sounds obvious is still a claim, and no census or state record was consulted, so the file names the sequence it can document and stops there.
- How much the tornado actually changed the pace, as distinct from the funding response to it. EPA's release is explicit that the money was to "expedite" the buyout, which establishes intent and a federal action. It does not establish how many households left sooner than they otherwise would have, and no source read here measures that.
- Picher's legal status and dates of dissolution. EPA states that Picher and Cardin "were dissolved" and are "currently uninhabited" as of the November 2014 document, but gives no dates or mechanism, and no Oklahoma municipal record was consulted. No dissolution date is asserted here.
- Present-day conditions. The most recent source read dates from November 2014. Cleanup at OU4 was described as ongoing over roughly a thirty-year horizon, so figures for tonnage removed, acres reclaimed and remaining work are certainly out of date, and are quoted with their date attached.
Sources
Every source below was read in full before it was cited. Archive snapshots are recorded at research time, because a file whose sources have evaporated is indistinguishable from one that never had any.
S1Tier 1Read in full
Tar Creek (Ottawa County) — Superfund Site Profile
US Environmental Protection Agency
EPA's own site record. Establishes the NPL proposal and listing dates, the 1979 mine-water discharge, the 1994 Indian Health Service blood-lead finding of approximately 35 percent above 10 µg/dL, the February 2008 OU4 Record of Decision, the role of the Lead Impacted Communities Relocation Assistance Trust, and the buyout totals of 628 residences, 74 businesses and 125 renters. It does not mention subsidence or the Army Corps at all, which is itself part of why the compressed version of this story survives.
S2Tier 1Read in full
Tar Creek Superfund Site, Ottawa County, Oklahoma — Community Involvement Plan Update, November 2014
US Environmental Protection Agency, Region 6 · 2014-11
The fifty-three-page EPA Region 6 document (SEMS 9699594), read in full including its appendices. It is the source of every load-bearing quotation in this file: the Army Corps study finding "a third or more of the homes in the area have varying degrees of collapse potential"; the public trust's voluntary relocation "for properties in Cardin, Picher, and Hockerville that are impacted by potential subsidence" on May 2006 legislative conditions; the May 2004 relocation limited to families with children under seven; the May 2003 interagency MOU naming subsidence as a physical safety concern; the 1996 blood-lead percentages for Picher, Cardin and Quapaw; the statement that Picher and Cardin "were dissolved"; the demolition start date of 24 January 2011 and the ODEQ closeout of 23 November 2011; the OU4 ROD and LICRAT; and the blood-lead reduction figures.
S3Tier 1Read in full
Panoramic view of a chat pile on the St. Joe property, Picher, Oklahoma — USGS media library item
S. Jerrod Smith · US Geological Survey, Oklahoma-Texas Water Science Center · 2013
The image's rights record and caption. Establishes the item-level "Sources/Usage: Public Domain." statement, the photographer's name and printed affiliation "US Geological Survey", the publishing science centre, and the caption "Panoramic view of a chat pile on the St. Joe property, Picher, Oklahoma, looking east." Read twice, the second time for literal on-page text rather than a summary.
S4Tier 1Read in full
NOAA National Weather Service, Tulsa forecast office · 2008-05-10
The weather service's own event page for this tornado, and the only source cited here for it. Establishes the rating and wind estimate — "EF4 on the Enhanced Fujita Scale with winds estimated at 165 to 175 mph" — the six fatalities, the northeast Oklahoma track of "approximately 29 miles, with a maximum width of one mile", the description of "extensive and widespread damage" to residential structures, and the 5:38pm radar timestamp as the tornado moved into Picher. It says nothing about the buyout, the Superfund site, or the town's population, and nothing of that kind is drawn from it.
S5Tier 1Read in full
EPA Provides Funds To Expedite Buyout of Picher Residents — news release, 21 May 2008
US Environmental Protection Agency · 2008-05-21
The document that connects the tornado to the pace of the relocation, in the agency's own words rather than by inference. Establishes the $8 million allocation to the Oklahoma Department of Environment Quality, split as $3 million directed by Congress for relocation of Picher, Cardin and Hockerville residents and $5 million through Superfund; the stated purpose of expediting "the buyout and relocation of residents"; the description of Picher as "heavily damaged by a May 10 tornado"; the deployment of emergency response staff for air monitoring and soil sampling; and the finding of "no health risks to residents or responders from the debris created by the tornado." Its headline verb is the reason this file can say "hastened" rather than merely placing two events near each other in time.
Checked for circularity: S1 and S3; S2 and S3; S4 and S1; S4 and S2; S4 and S3; S5 and S3; S5 and S4 were confirmed not to derive from one another. Two sources that trace back to the same origin are one source, and counting them twice is the most common way a false claim looks well-evidenced.
How we checked this
What we went looking for, and what we found
Five checks. Two changed the file, and one of them changed it after the file was first written.
First, against overcorrecting. The obvious move once the subsidence finding turns up is to say "it wasn't the lead after all", and that would be worse than the claim it replaces. Lead is why the site was listed in 1983, why EPA remediated roughly 2,887 yards and public properties, and why the later buyout ran under a trust named for lead-impacted communities. The verdict was written to keep both, and a contradicting-evidence entry states plainly that a reader who concluded lead was incidental would be more wrong than one who thought it was everything.
Second, against merging the two relocations. There is a 2006 state subsidence trust and a 2008 CERCLA OU4 relocation implemented by LICRAT, and it is tempting to treat them as one programme with one motive. EPA names LICRAT only for the second. Recorded as an uncertainty rather than resolved by assumption.
Third, against the blood-lead number. Two EPA documents give different figures — approximately 35 percent of Indian children tested in 1994, and 38.3 percent of children tested in Picher in August–September 1996. They are different populations and programmes. Both are quoted with their scope attached rather than picking the higher one for the card.
Fourth, on the Corps study itself, a documented limit rather than a finding. EPA reports the conclusion and the requesting senator but does not date or name the study, and the study was not retrieved. The file quotes EPA's characterisation as EPA's characterisation and asserts nothing about the Corps' methodology.
Fifth, and against this file's own first version, which failed the same test it applies. The original headline read "Mine collapse is what emptied it". That substitutes one single cause for another and is exactly the error the file exists to correct. Rechecked against the record, the documents describe a sequence: lead drove the 2004 programme for families with children under seven; the Corps study widened the buyout to homes at risk of collapse from May 2006; and an EF4 tornado on 10 May 2008 was followed eleven days later by EPA funding to "expedite the buyout".
Two sources were added to support the third element rather than asserting it from general knowledge — the NWS Tulsa event page for the tornado itself, and the EPA news release of 21 May 2008 whose headline verb is "Expedite". Before those were located, the EPA community involvement plan cited throughout this file was searched for "tornado" and returned zero occurrences, which is why the claim could not have been made on the original source set and was not.
A fourth element offered alongside these — that Picher had been losing population for decades as mining wound down — was not adopted, because no source read here gives a population figure for Picher in any year. It is recorded as an uncertainty instead.
- We searched for evidence against our own conclusion
- We traced where the claim started
- The claim is stated in its strongest form, not a weak version
- The photograph was verified independently of the story
- We checked whether this was settled after the popular version froze
- Image rights were traced to a stated licence
- We checked what we were leaving out
- The verdict sat before publication, and was re-read cold
- Confidence in the verdict
- high
- Verdict assigned
- Most recent authoritative source
- Researched by
- The Strange File
REVISED 2026-07-31, AND THE COOLING-OFF CLOCK WAS RESET RATHER THAN ALLOWED TO MATURE. The original verdict was assigned 2026-07-30, so by the calendar the cooling-off gate would have passed today on its own. It has instead been reset to 2026-07-31, because the verdict itself materially changed today: it went from naming a single cause to describing a three-part sequence, and two sources were added to carry the third part. Cooling off is a wait on a verdict, so a verdict rewritten today has not waited. Letting yesterday's date clear today's reasoning would be satisfying the control on a technicality, which is the failure mode D-053 exists to prevent.
COOLING-OFF PASSED 2026-08-01. The verdict was assigned 2026-07-31 and the wait has elapsed. The fresh-eyes pass was made by a session that neither researched nor wrote this file, re-reading the dossier, the evidence, the sources, the verdict reasoning, the rights record and the Quick File cold. Nothing required changing. Every specific in the headline and hook traces to quoted source text: the May 2004 programme covering "families with children under the age of seven", the Corps study finding "a third or more of the homes" with collapse potential, and a tornado that struck while the buyout was already running. The headline states a sequence and the record describes a sequence.
That pass is cooling-off and nothing else. Cooling-off, independent review and the owner's approval remain three separate controls and none may stand in for another. No independent review has occurred; this file carries no sensitivity flag, so none is required. The owner has not approved publication, and READY is not permission to publish.
Sensitivity considered and not flagged, but it was a real question rather than a formality. This file concerns a community whose children were measurably harmed, and many of the people relocated are living. The subject here is the documentary record of two government programmes, not any individual's health, and no resident is named, quoted, or depicted — the photograph contains no people and no identifiable property. Named individuals are limited to a US senator acting in office and a USGS photographer credited on his own image. Nothing in the file makes a claim about any private person.
The file also deliberately avoids the register the subject invites. "America's most toxic town" is how this place is usually introduced, and it is quoted once, as the circulating form, rather than adopted.
Part of a collection
What The Stamp Actually CertifiesA grant, a listing, a prize. Each certifies something narrower than everyone reads it to mean.
The Correction OvershootsDebunking has a second error, and almost nobody guards against it.
Places America LeftTowns emptied by industry, and the sequence the retelling flattens.
Also related
The USGS calls 1905 the Salton Sea's most recent formation, not its first
The Colorado broke through a half-finished irrigation headgate in 1905 and took until 1907 to put back. But the river had filled that basin on its own before, and each time the basin dried out again.
Kīlauea makes hair out of glass, and it blows into people's gutters
Strands of volcanic glass over a metre long and under a millimetre thick, spun from lava the way sticky dough pulls into threads. The USGS advice for finding some is to leave it alone.