File / colorado-springs-2022
The Object over Cheyenne Mountain
Five soldiers, accounts written down years later, and an explanation that depends on which time of day you believe
In February 2022 five U.S. Army soldiers from one unit saw a white object hold still over Cheyenne Mountain, Colorado, and then vanish. The file holds two FBI interviews taken two and three years later and an undated analysis that attributes the sighting, with low confidence, to sunlight reflected off snow onto low cloud. No photograph was taken, the five drawings and the FBI artist's sketch are not in the file, and the accounts disagree on the time of day, the duration and what the object looked like.
Official position
The only assessment in the file is an analysis whose author, addressee and date are redacted. It judges that the object "was possibly backscattering of sunlight" reflected off snow-covered ground onto low-level cloud, states that "no anomalous data or characteristics were recorded or assessed, and the event did not represent an unknown adversarial capability", and holds the assessment with low confidence because of uncertainty in each witness's field of view, the snow cover, and the elevation and amount of cloud. It adds that no aircraft or balloons were noted active in or around Cheyenne Mountain at the time.
Alternative hypotheses
- Sunlight reflected off snow onto low cloud. The analysis's own hypothesis, held with low confidence. It rests on a mid-morning sun in the south-east, which fits the analysis's 09:35 and not the early afternoon given in the FBI interview of March 2025, and it needs cloud the witnesses say was not there.
- A balloon or a blimp. The soldiers considered both at the time and set drift aside because the object held its place against the saddle of the mountain for minutes. The analysis says no aircraft or balloons were noted active at the time.
- An unidentified object. The accounts describe a stationary white object with a surface of lines or shifting panels that vanished in the time it took to look away. They rest on eyesight alone, written down two and three years later, with no image and no instrument.
What the three documents describe
Three documents describe a sighting over Cheyenne Mountain, Colorado, in February 2022, and they share a core.
Five soldiers from one U.S. Army unit are outdoors, looking west. Above the mountain hangs a white object that does not move. It stays for somewhere between half a minute and five minutes, and then it is gone, suddenly, without being seen to leave.
In June 2024 an FBI special agent and a forensic artist interviewed one of the five, “a former U.S. Army intelligence officer”, in New York (FD-1057). In March 2025 two FBI agents interviewed one of the five by Microsoft Teams (FD-302). A third document, an analysis whose author, addressee and date are all redacted, works from “the accounts of five U.S. Army service members” and offers one explanation.
The vocabulary overlaps. The June 2024 interview calls the object “potato” shaped, with “distinct edges”, “somewhat translucent with a slight shimmer”, made of “fish scales or panels”. The analysis calls it “an angular, non-symmetrical potato made of uneven panels” with “defined edges” and “a slightly translucent, shimmering white appearance”. The March 2025 interview has its own words for something similar: “bean-shaped”, “matte white, or off-white”, covered in “intersecting lines or ridges which formed an abstract polygon pattern”.
Whether it is one event
Nothing in the file says it is. The file number of the FD-302 and the case ID of the FD-1057 are redacted. Neither FBI document mentions the other, and the analysis does not say where its five accounts came from or whether either interview was among them. The June 2024 interview redacts even the year of the sighting: it reads “On February”, and the rest of the date is blacked out.
The circumstantial case is strong: the same mountain, the same month, five witnesses from one unit in every account, an object that held still and then vanished, and descriptions that share words. It is still circumstantial. This page treats the three documents as probably describing one sighting, and does not present that as established.
Where the accounts part
Set side by side, without smoothing:
| FD-1057, June 2024 | FD-302, March 2025 | Analysis, undated | |
|---|---|---|---|
| Date | “On February” and a redaction; the year is not visible | “a workday in early February 2022” | 15 February 2022 |
| Time | “approximately 11:25 am” | “approximately 1:00 or 2:00PM in the afternoon” | “0935 MNT”; the sun is computed for “0945 MST” |
| Duration | “approximately two minutes” | “approximately 3-5 minutes” | “approximately 30 - 180 seconds” |
| Appearance | “potato” shaped, “creamy/whitish opalescent”, “somewhat translucent”, panels that “shifted in slow waves” on a still object | “matte white, or off-white”, “appearing non-metallic”, “bean-shaped”, “completely motionless”, covered in lines or ridges | “angular, non-symmetrical potato made of uneven panels”, “completely stationary” while “slowly changing shape”, “slightly translucent” |
| Size | not given | “10% of the saddle of the mountains, or about the size of a pool cue at arms-length” | “roughly the size of a large jet” |
| Disappearance | “cloaked” in the time “it took to turn a head” | they all looked away to send for a phone; “When they looked back the object had vanished” | “despite keeping their eyes on the object, it suddenly disappeared” |
Three times of day, from mid-morning to the early afternoon, and two incompatible accounts of the vanishing: one in which everyone looked away, one in which nobody did.
The one explanation in the file, and what it rests on
The analysis offers a single hypothesis. Sunlight reflecting off snow on Cheyenne Mountain lights low cloud nearby, which “might account for the visibility of the object followed by its sudden disappearance” if the cloud or the sun “shifted slightly”. It holds that with low confidence.
The hypothesis is built on a position of the sun, and the analysis states which one: at 09:45 MST on 15 February 2022 the sun stood about 27.5 degrees above the horizon, at an azimuth of about 125, “in the southeast sky”. That is a mid-morning sun in the south-east. It fits the 09:35 the analysis gives for the sighting. It does not fit the March 2025 FBI interview, which puts the sighting at “approximately 1:00 or 2:00PM in the afternoon”, and the analysis does not test the June 2024 interview’s “approximately 11:25 am” either. If the afternoon time is right, the geometry the explanation rests on is not the geometry of the sighting.
The analysis is not consistent about its own clock. It dates the sighting to “0935 MNT” and computes the sun for “0945 MST”: two times, and two names for the time zone.
It also records a conflict it leaves open. The witnesses describe a clear sky, “no clouds” in June 2024 and “clear blue skies” in March 2025, while the weather reports the analysis cites, from AFWA and Weather Underground, suggest the sky was “partly to mostly cloudy that morning”. The hypothesis needs cloud. The witnesses say there was none.
Drawings that are not in the file
Every account says the object was drawn, and no drawing is in the file. The March 2025 interview says each of the five “independently drew what they remembered the object to look like”, compared the drawings afterwards, and found them “consistent in showing a horizontal, white, ‘bean-shaped’ object with abstract lines crossing on it”. The June 2024 interview encloses a “Sketch draft” by the FBI forensic artist and says further drafts would follow “in order to complete the final sketch product”. The release contains neither the five drawings nor any sketch.
What it does contain is a single image, FBI-UAP-D003, published as a “Digital Rendering”: a white oval object in a blue sky above a photograph of a mountain, with houses and a road barrier in the foreground. It carries no author, no date and no statement of what it was based on. The image inside the file has no metadata of its own, and the only dates attached are those of the file prepared for publication in June 2026. It cannot be assigned to either interview, and nothing in the file says it is the artist’s sketch.
Two observations, recorded as observations and nothing more. The object in the image is horizontal, white and covered in a network of polygons, which is closer to the “bean-shaped” object “with abstract lines crossing on it” of the March 2025 interview than to the shifting “fish scales” of June 2024. And no snow is visible anywhere in the image, in the foreground or on the mountain. The analysis’s 6 to 12 inches of snow refer to the mountain itself; since nothing records when the background was photographed, that is a note, not a contradiction.
Contacted, not interviewed
The March 2025 witness says AARO contacted him in early 2024, that he provided a response, and that he “was never interviewed by AARO or anyone else”. The June 2024 witness was interviewed in person by the FBI, with a forensic artist, nine months before that statement was recorded.
Either the two interviews are with two different soldiers of the five, or one of the accounts is imprecise about who has interviewed whom. The file does not say which, and this page does not decide. It matters: two witnesses would be two accounts from one group; one witness twice would be one account told two different ways.
Why this scores where it does
Witnesses score 3: five soldiers from one unit, one of them described as a former Army intelligence officer, all seeing the same thing at the same moment. That is more than a single observer and less than it sounds, because they were standing together, and the only first-hand accounts in the file were taken two and three years later.
Multi-channel scores 1. Five people in one place are one channel, and it is the only one. Radar, optics and physical trace are all 0: no radar, no camera, no instrument of any kind. Five witnesses, and none of them had a phone; by the time one was to be fetched, the object was gone.
Documentation scores 3: two FBI interviews and an analysis with its reasoning exposed where it is not redacted, against an analysis with no author, addressee or date, a year of the sighting that one interview redacts, and a time of day on which no two documents agree. Anomaly scores 2: an object that holds still and vanishes is unusual as reported, but the report is all there is, and the file carries a conventional explanation it never tested against the other times. Data today scores 2: the place and the month are known and the analysis names its weather sources, but without the time of day nothing can be recomputed with confidence.
That puts the case at 1.3 overall, class D, testimony only, with the status insufficient data. Not because five soldiers are not credible, but because credibility is all the file holds.
What would settle it
The time of day. 09:35, 11:25 or the early afternoon. The one explanation in the file depends on it. The March 2025 interview says one of the five made “an official report to the army about their observation”; that report would be the nearest record to the event, and it is not in the file.
The redacted pages of the analysis. All of page 2, and the blocks at the top of page 3 and the foot of page 4. What they contain is unknown. If they name the author, the date or where the five accounts came from, they would settle whether the FBI interviews were among them.
The drawings. The five sketches the soldiers made and compared, and the FBI artist’s draft and final sketch. They are the nearest thing to an image of the object that exists, and they would show whether the five saw one shape or two.
← All case files