The Paintsville Train Collision
The case exists as a single first-person account submitted anonymously to the National UFO Reporting Center. This case file covers what witnesses reported, the official narrative, and a two-pass assessment with its evidence tier.
What did witnesses see at Paintsville?
The case exists as a single first-person account submitted anonymously to the National UFO Reporting Center. The author identifies himself as the engineer of a CSX coal train. In his words, "At exactly 2:47 a.m. on Jan. 14 02, while working a coal train enroute from Russell Kentucky to Shelbiana Kentucky, our trailing unit and first two cars were severely damaged as we struck an unknown floating or hovering object." He says he knows the precise time because his watch froze at 2:47 and "to this day shows that time," and that the electrical systems on both locomotives "went haywire" at the same moment.
He describes approaching a bend near milepost 42 in a stretch the crews called "the Wild Kingdom" for the animals seen there. He lays out the geography as the river, the #1 track, the #2 track, and "a straight up mountainside, carved out for the laying of these tracks." Seeing lights ahead, he assumed an oncoming train and killed his own headlight so as not to blind the other crew. As they rounded the corner the onboard computer flashed in and out, the speed recorder "went nuts," and both locomotives died. Alarm bells rang, and that is when he says he and his conductor saw the objects.
There were at least three, he wrote, "apparently scanning the river for something," each with several "search" lights trained on the water. The nearest hovered ten to twelve feet above the track. He described it as metallic silver "with multiple colored lights near the bottom and in the middle," no windows or openings, roughly 18 to 20 feet long and ten feet high. With both engines dead the train coasted nearly silently, and he says "the first object did not respond in time." He estimates they struck it at 30 mph with 16,000 trailing tons behind them. The object "clipped the top of our lead unit then skipped back slicing a chunk out of our trailing unit and first two coal cars." The other objects, he wrote, "vanished into thin air."
The emergency brakes set from the loss of power and the train stopped a mile and a half to two miles later. Power restored once they were stopped, and they notified their dispatcher "located in Jacksonville Fla." He says the two damaged coal cars "looked as if they'd been hit with a giant hammer." He limped the train into the old Paintsville yard at about 5:15 a.m., where the yard's overhead lights were dark and the only light came from official-looking vehicles. The rest of the account describes an aftermath rather than a sighting: a crowd of men in "weird outfits," no railroad officials anywhere, a leader who introduced himself as "Ferguson," hundreds of questions, a confiscated cell phone, the two locomotives and two cars towed off and parked under a large tent "four tracks over." The crew was told that "due to national security our silence on this matter would be appreciated," driven to Martin, Kentucky for more questioning and drug testing, then sent on to Shelbiana to rest. Working back through Paintsville hours later, he wrote, there was "no sign of the engines, cars, tent, people, nothing."
What is the official explanation?
There is no official narrative because, on the available record, no official body ever opened a file. That absence is itself the central problem. A 16,000-ton train striking an object hard enough to crush two coal cars and demolish a locomotive cab is, under federal law, a reportable rail equipment accident. The Federal Railroad Administration requires carriers to file damage and accident reports under 49 CFR Part 225, and those reports populate a public FRA accident database. No FRA accident entry, no NTSB action, and no CSX incident record corresponding to a destroyed locomotive and two wrecked cars at Paintsville on 14 January 2002 has ever been produced by anyone, including the witness.
The closest thing to an official comment sits in NUFORC's own editor's note on the report. NUFORC director Peter Davenport, who runs the database and first circulated the account, flagged the report as one that "may be a hoax." NUFORC's note records that the J. Allen Hynek Center for UFO Studies (CUFOS) looked into it on follow-up, and that a former CSX railroad official stated no such incident was documented during his tenure. So the only institutional voices that touched this case, the reporting center that published it and a retired carrier official, both came down on the side that it did not happen as described.
The one detail in the account that checks out against real CSX operations is small: CSX does dispatch from Jacksonville, Florida, which the witness named correctly. That single accurate fact is the high-water mark of the official record. Everything else that would normally generate a paper trail, a federally mandated accident report, an insurance claim on destroyed equipment, a locomotive written off the active roster, a union grievance over a crew detained and drug-tested by unidentified men, leaves no trace in any official archive.
What did the witnesses think it was?
The case has, in effect, one witness and one document. The submitter says he was the engineer and that his conductor was beside him, but the conductor never came forward, was never named, and never corroborated anything. NUFORC accepted the report anonymously, so there is no name to weigh, no employment record to confirm, and no way to test whether the author ever worked for CSX. The witness clearly believed, or wished readers to believe, that the crew had collided with a genuine craft and that a government apparatus erased the evidence overnight under a national-security pretext.
One later NUFORC report is sometimes offered as corroboration. In a submission posted around 2010, an anonymous person who said they were a Paintsville firefighter and paramedic described driving home after a shift late on 13 January 2002, hours before the claimed collision, and seeing a "big light zig-zag" across the road, then two more lights flanking the car, while the vehicle's gauges "went crazy" and the lights flickered. That witness opened with "don't know if train/UFO was (a) hoax but do know there was something strange in the area that night." It is thin, also anonymous, and was filed years after the fact, so it corroborates a mood more than a collision. It cannot independently establish that a train hit anything.
Against the witness sit people who actually work the territory. On railroad forums where the report circulated from 2006 onward, CSX-area railroaders picked the geography and physics apart. They noted that a loaded coal train would normally run down toward Russell rather than away from it over that route, identified the waterway correctly as the Levisa Fork of the Big Sandy River, and questioned the dynamics: one poster, Paul D. North Jr., calculated that an emergency stop in "a mile and a half" from 30 mph implies a deceleration of about 0.17 mph per second, far gentler than a real 16,000-ton emergency application, which would stop the train much sooner. A railroader posting as beaulieu offered a mundane alternative for the locomotive damage, that it "is consistent with what happens when you hit the loading spout at a flood loading tipple." A self-identified retired CSX engineer of 33 years on the Cincinnati-Corbin Division pointed out that two locomotives and two cars could not have been spirited out of Paintsville yard overnight by truck, "they had to move by rail," undercutting the vanishing-equipment finale. These are motivated skeptics in the sense that railroaders defend their trade's plausibility, but their objections are concrete and checkable, not character attacks.
The dispute
The dispute holds that the Paintsville incident never physically happened and that the single anonymous account, an unnamed CSX engineer describing his coal train striking a hovering metallic object near milepost 42 at 2:47 a.m. on January 14, 2002, reads as fiction rather than a record of an event. The core counter-explanation is twofold: there is a total absence of the records such a collision would mandatorily generate, and the physics and logistics described do not survive scrutiny. No Federal Railroad Administration accident report under 49 CFR Part 225, NTSB action, or CSX incident record matches the event. NUFORC director Peter Davenport flagged the submission as one that "may be a hoax," and when the J. Allen Hynek Center for UFO Studies looked into it, a former CSX official said no such incident was documented during his tenure. Because the engineer never identified himself, the named conductor beside him never came forward, and a later "firefighter" account is itself anonymous and only describes lights hours earlier, the entire case rests on one uncorroborated submission.
What lifts this well above a bare official denial is that civilian analysts showed their work. Paul D. North Jr. reconstructed the braking and found the implied deceleration was about 0.17 mph per second, far gentler than a genuine 16,000-ton emergency application would produce, meaning the narrative's own numbers contradict a real emergency stop. A railroader posting as beaulieu matched the described locomotive damage to a mundane cause, that it "is consistent with what happens when you hit the loading spout at a flood loading tipple," rather than to a midair impact. A self-identified retired CSX engineer of 33 years on the Cincinnati-Corbin Division pointed out that two engines and two coal cars "had to move by rail" and could not have been spirited out of Paintsville yard overnight by truck, so the claimed wreckage could not have vanished from the record as the account requires. Skeptics further noted that the loaded train's direction contradicts typical Russell-to-Shelbiana routing, that classic paranormal tropes such as a stopped watch frozen at 2:47 and figures in "weird outfits" appear, and that the one operational detail offered, the Jacksonville dispatch reference, is public knowledge any railfan would know. This is a method-shown debunk, not a bare apparatus assertion: a candidate alternative cause for the damage, a calculated physics mismatch, and a demonstrated logistics impossibility all point the same way.
Where it falls short of closing the case is that the convergence is built from absence and probability, not from a closure event. There is no confession or recantation, because the author was never identified and never took the story back. There is no recovered prop or rigged photograph to tie to this event, because the case is one text document with nothing physical attached. And there is no positive identification of what actually happened, because "a tipple spout could do that" is a proposed alternative, not a demonstrated one, and missing FRA paperwork proves the absence of a filing, not the presence of a hoaxer. The braking math and the towing point are strong, but they argue that the event almost certainly did not occur, which is a near-certainty rather than the proven fabrication this archive requires before it will say a core claim has been shut. So the counter-case stands at full strength and the page records it in full, while the central claim, uncorroborated as it is, has not been positively closed, which is why it holds at Disputed rather than moving to a discredit verdict.
Is the Paintsville Train Collision real? The two-pass assessment
Pass one, how this is entirely ordinary. The simplest reading is that the report is a piece of creative writing, a campfire story dressed in railroad jargon and filed anonymously roughly four years after its supposed date. Everything that would anchor it to reality is missing and conveniently explained away inside the story itself: the equipment was towed off and hidden under a tent, the witnesses were sworn to silence, the crew was anonymous, and the whole site was cleared before sunrise. That is the shape of an unfalsifiable tale, every avenue of verification closed by the narrative. The hard facts cut against it. A collision that crushed two coal cars and demolished a cab would be a federally reportable FRA accident under 49 CFR Part 225, and no such record exists. The braking physics that named analysts worked through do not match a real 16,000-ton emergency stop. Paul D. North Jr. reconstructed the implied deceleration at about 0.17 mph per second, far gentler than a genuine emergency application. The locomotive damage has a mundane match in striking a loading spout at a flood loading tipple, per a railroader posting as beaulieu. Two engines and two cars cannot vanish overnight by truck from a rail yard, per a self-identified 33-year retired CSX engineer who noted "they had to move by rail." The route is backward for a loaded coal train. The stopped watch at 2:47 and the men in "weird outfits" are stock paranormal furniture. The lone correct detail, Jacksonville dispatch, is public knowledge any railfan would know. Taken together this reads as a hoax or a tall tale, and the reporting center that published it, NUFORC under Peter Davenport, flagged it as one that "may be a hoax," with CUFOS follow-up and a former CSX official confirming no such incident was documented during his tenure.
Pass two, what the debunk does not close. Strip the case to its load-bearing claim and notice what the skeptical case is actually made of: an absence of records, an internal physics mismatch, and a speculative mundane alternative. None of those three is a demonstrated account of what actually happened on that track. The author was never identified and never recanted, so there is no confession. There is no physical prop, model, or rigged photograph to recover and expose, because the case is a single text document with no image attached. And no investigator ever produced a positive identification of a real object or a real event that the engineer mistook or fictionalized. The railroaders showed real method, but "the numbers imply this almost certainly did not happen" is a probability judgment, not a closure. The corroborating 2010 NUFORC report from an anonymous self-described Paintsville firefighter, who saw a "big light zig-zag" with gauges that "went crazy" on the night of 13 January 2002 and opened with "don't know if train/UFO was (a) hoax but do know there was something strange in the area that night," is thin and proves nothing on its own, but it has also never been retracted or explained away. What survives, then, is a structured-craft account with electromagnetic effects that no one has affirmatively unmasked.
Tier and why not Strongly Disputed. Strongly Disputed is reserved for a core claim that has been positively closed: a confession or recantation, a recovered prop tied to this event, or a positive identification of the specific cause. This case has none of the three. What it has is strong, civilian, method-shown skepticism, the braking math, the route logic, the towing impossibility, a candidate mundane damage cause, and the publishing center's own "may be a hoax" verdict, that all point the same way without ever landing the closing blow. That is exactly the top of the disputed band, where the counter-case is serious but argues absence and probability rather than proven fabrication. So it holds at Barely Disputed. The honest position is that this almost certainly did not happen, but on this archive's method "almost certainly" is not "demonstrably," and the final step to a discredit verdict would need a confession, an exposed fabrication, or an identified real cause that no one has yet produced.
Sources
- www.nuforc.org/webreports/049/S49934.html
- web.archive.org/web/20250514022755/https://forum.trains.com/t/update-2002-paintsville-kentucky-collision-with-unidentified-object/195099
- web.archive.org/web/20251207012030/https://forum.trains.com/t/railroaders-ufo-collision-csx/133322
- www.ufoinsight.com/ufos/close-encounters/kentucky-ufo-train-collision
- www.trainorders.com/discussion/read.php?2%2C1205108
- www.ecfr.gov/current/title-49/subtitle-B/chapter-II/part-225
- data.transportation.gov/Railroads/Rail-Equipment-Accidents-Accident-Detail-Report/e5vm-43rk
More cases from this region: UFO sightings in United States
