A Christmas Letter From Budapest: The Saucer Report the CIA Filed Twice

In November 1955, a naturalized US citizen received a letter from his niece in Budapest, and it ended up in CIA channels. Between Christmas shopping and a visiting German circus, she described "so-called saucers" that had kept Hungary's people in a nervous state and its scientists busy, with a speed estimate of 12,000 kilometers per hour. The CIA typed it up as a formal Information Report, restricted it from foreign eyes, and kept a photostat of the original in the CIA Library. The 2026 PURSUE release contains that report twice, under two different file IDs, plus a third Budapest file whose text extraction this site cannot responsibly quote.

What these files are - and why two of them are the same report

The PURSUE release catalogs three CIA files connected to UFO sightings in Hungary: CIA-UAP-009 ("Unknown Flying Objects Observed Over Budapest," dated 1957), CIA-UAP-013, and CIA-UAP-018 (both titled "Report of Unusual Flying Object Sightings and Attendant Scientific Activity," cataloged under 1956 and 1955 respectively). On paper, that looks like three separate Cold War sighting reports from behind the Iron Curtain. The extracted text tells a different story, and this page follows the text.

The plain-text extractions of CIA-UAP-013 and CIA-UAP-018 carry the same CIA report number - 00-B-93674 - the same subject line, the same source description, and the same four numbered paragraphs, sentence for sentence. (In the raw OCR the number renders as "00- B--93674" in one copy and "00- B--9.3674" in the other; the digits match.) On the text alone, these are two copies of the same one-page CIA Information Report. They are, however, two genuinely distinct PDF files: 236,547 and 277,577 bytes (their file pages display these as 231.0 KB and 271.1 KB), with different SHA-256 hashes, and the CIA-UAP-018 file contains something the other copy's extraction does not - an additional page carrying a fuller English translation of the underlying letter, nearly all of it cleanly readable.

Why the same report appears under two file IDs is not something this site can answer from the data, and we will not guess. The release metadata for both files notes that a more redacted version "has been available on CIA's public website" (CIA-UAP-013's summary calls it "this report", CIA-UAP-018's "the report"), which allows for the possibility that the two PDFs are different scan or redaction generations - but the plain-text extraction cannot settle that, so this page documents the overlap and stops there. The metadata also dates the two files a year apart (1956 and 1955); the report's own first paragraph, in both copies, dates the letter's receipt to November 1955.

What is not in dispute is what the report is. The form header reads "PREPARED AND DISSEMINATED BY CENTRAL INTELLIGENCE AGENCY", and stamped across the middle of the page, cleanly legible in the extraction, is the CIA's standard caveat:

"THIS IS UNEVALUATED INFORMATION"

Verbatim from the CIA-UAP-013 extraction, page 1. The CIA-UAP-018 copy renders the same stamp with an OCR typo ("IMFORMATION"). The header line is normalized from "PREPARED ANO DISSEMINATED .BY / CENTRAL INTELLIGENCE AG.ENCY".

That caveat matters for everything that follows. The CIA was not certifying that saucers flew over Budapest. It was recording, without judgment, what a source told it - which is exactly what makes the document interesting seventy years later. It shows what kind of information the Agency considered worth formalizing, restricting, and archiving at the height of the early Cold War.

The source: an uncle in America, two nieces in Budapest

The source line on the report's cover describes the reporting person - not the witness - in one sentence. OCR damage obscures a few words, but the readable portion (normalized) describes a "Naturalized US citizen of [word garbled] extraction" who "corresponds regularly with two nieces living in [word garbled]." The release metadata fills the gaps the OCR leaves: the information "came from a letter between Hungarian relatives, living in the USA and Budapest." The report's first paragraph confirms the city in the source's own words: "In November 1955 I received a letter from my niece in Budapest."

The source line is normalized from the CIA-UAP-018 extraction ("Naturaliud OS· citis~n:f:': extraction. He correspoma regµJ,arl;y nth two m.eces liVi.llS 1n t."). The paragraph-1 sentence is normalized from that copy's "In l'lOirEIIIJlle'r 19,55· I received f!. le~ .frGIII ~ niece in ·~est." - the words "November" and "Budapest" are OCR-damaged in both copies but recoverable; "letter" is damaged only in this copy (the CIA-UAP-013 copy renders it cleanly); "niece" is clean in both copies, with only the preceding "my" damaged. The location is independently anchored by the letter's own text and the release metadata.

Stop on that arrangement for a second, because it is the whole story in miniature. In 1955, Hungary sat behind the Iron Curtain under a hard-line communist government; the Hungarian Revolution was still a year away. The CIA had no officers strolling Budapest collecting saucer reports. What it had was émigrés - naturalized citizens whose family mail crossed the Curtain - and a standing apparatus for turning that mail into intelligence. A niece's chatty family letter, mentioning something strange in the sky, was collectible foreign intelligence. The CIA assigned it a report number, and the Espionage Act boilerplate on the form (citing sections "793 and 794" of Title 18) framed a family Christmas letter as national-defense information.

The letter: Christmas trees, a German circus, and saucers

The second page of the CIA-UAP-018 file is the reason this page exists. It carries an English translation of the niece's letter, and unlike the report cover, the OCR read it nearly clean. It opens with a line that lands differently when you remember where and when it was written:

"I don't know whether the letter I sent before will arrive. I wish you all a very Merry Christmas and a happy New Year. Here, the children are preparing for Christmas. The shops are full of toys and Christmas trees. The streets are crowded with window shoppers."

From the CIA-UAP-018 extraction, page 2. OCR normalizations: "J:r.nm;r" rendered as "know"; "rhe shops" as "The shops"; a stray mark after "trees" removed. All other wording verbatim.

The letter is almost entirely ordinary life. The children's vacation runs "from 20 December through 17 January". Someone "took a day off on All Souls' Day". The writer asks her uncle, "you write so little. How are you? Has the family grown? Are you all well?" A relative, name redacted, "is having a lot of trouble with her back"; another "has been sent off for a week to do some work in the province, he won't return till the evening of the 24th." And then, deadpan among the family news:

"A famous German circus is on tour here now with many animals and artists. They say it is very good. We are planning to see it with the children."

From the CIA-UAP-018 extraction, page 2. OCR normalizations: a doubled period after "artists" and a trailing mark on "children" removed. The other lines quoted above carry minor disclosed repairs: "d~y" rendered "day"; "How are your·" and "grown7" rendered with question marks. In the unquoted framing, the raw "fhe childre~" is rendered "the children." Redaction boxes sit before "is having a lot of trouble" and "has been sent off for a week", and between "All Souls' Day" and "you write so little".

Two sentences later comes the paragraph the CIA filed the letter for:

"Everyone has been excited by the so-called saucers for the past few weeks. These fast-rushing heavenly phenomena have been and still are keeping scores of scientists busy - [redaction] has surely seen them or read about them; these amazing fliers moved at a speed of 12,000 kilometers per hour."

From the CIA-UAP-018 extraction, page 2. OCR normalizations: "Ev.eryone" rendered "Everyone"; "phenorna" rendered "phenomena"; "~liers" rendered "fliers"; a stray mark before "excited" (raw: "·excited") removed. A blank redaction box sits mid-sentence before "has surely seen" (the extraction renders the join as "-~I_____~ras"). All other wording verbatim, including "so-called saucers," "fast-rushing heavenly," "keeping scores of scientists busy," and "a speed of 12,000 kilometers per hour."

The letter closes the way family letters close: "We wish you all good health and send our regards and affection to mother and the children. Much love and many prayers," - that line is verbatim and needed no repair at all.

Nothing in the saucer paragraph is presented as the writer's own sighting. She reports a weeks-long public excitement, ongoing scientific attention, and a specific speed figure - 12,000 kilometers per hour, roughly 7,400 mph, on the order of ten times the speed of sound - that she attributes to what people had heard, not to anything she measured. No aircraft flying anywhere in 1955 approached that speed. The figure is an unattributed popular estimate relayed in a family letter, and this page treats it as exactly that. But the claim that the phenomenon was "keeping scores of scientists busy" in a Warsaw Pact capital, written by a person with no apparent motive to invent it for her uncle, is the detail a CIA reports officer would have underlined.

Two translations of the same passage

The report's cover page quotes the saucer passage in the source's own words - paragraph 1 flags it as a "free translation" (itself a repaired reading: the phrase is OCR-mangled in both copies, rendered "['tree t.ttlnslatioy'" in CIA-UAP-013 and "/Jree tftzlslati~" in CIA-UAP-018) - and the wording differs from the fuller translation above in ways worth seeing side by side. The free translation's readable portion includes a phrase that does not appear in the page-2 version:

"The so-called [flying] saucers (rockets) [sic] for several weeks kept the people in a nervous state."

Normalized from the CIA-UAP-018 extraction, page 1 (raw: '~e ao~ "f~ saucera (roc~ts) [rde7 for Se"reral wee~ kept the people in a nervous state.'). The clause "kept the people in a nervous state" is clean and verbatim; "saucers (rockets)" is lightly repaired from "saucera (roc~ts)"; "[flying]" is bracketed because the word is OCR-destroyed in both copies; the bracketed "[sic]" appears in the document itself (rendered "[sie7" in the CIA-UAP-013 copy). The corresponding sentence in CIA-UAP-013 is more heavily garbled but matches word for word where readable.

And the speed claim recurs, with the cleanest fragment reading:

"... km per hour was estimated on these ..."

Verbatim fragment from the CIA-UAP-013 extraction, page 1. The tokens immediately before and after (rendered "J:2 'tllowlawi" and "!.eyers" in the raw OCR) are consistent with "12 thousand" and "flyers" and with the attachment's clean "12,000 kilometers per hour," but are too damaged to quote as words, so we do not.

Note what the free translation adds: "(rockets)", with a "[sic]" - a parenthetical suggesting that in the letter-writer's mind, or the source's, the saucers and rockets were interchangeable candidates for what was speeding over Hungary. That is a small, honest window into 1955: two years before Sputnik, with Soviet missile development underway but almost entirely invisible to the public, an anomalous fast-moving light could be a saucer or a rocket, and a Budapest family letter did not bother to choose. The middle of the free-translation passage (a clause about the scientists) is too OCR-damaged in both copies to quote responsibly; its readable fragments are consistent with the attachment's "keeping scores of scientists busy," and we leave it at that.

The sketch: Budapest to Moscow

Paragraph 2 of the report records that the letter came with a drawing. The normalized sentence reads: "Inclosed with the letter was the following sketch indicating the formation and suspected course of the above objects" - and beneath it, in both copies, the extraction picks up exactly two place labels from the sketch itself: "Budapest" and "Moscow". The word "Moscow" is clean and verbatim in both files; "Budapest" is garbled in both ("Bu&tpest·" with a trailing mark in one, "Bwlapest" in the other) but unambiguous.

Paragraph-2 normalization from the CIA-UAP-018 extraction ("Inr~ ~tb the l.e'tter was the fo~ng sketch 1ndicat1ng the :t'onmstion m:i.d SU11,Pected course of the above obJects:"); the words "sketch," "course of the above," and "objects" (as "obJects") are recoverable; "formation" is repaired from ":t'onmstion" and corroborated by the CIA-UAP-013 copy's ":formation". The sketch itself is an image and does not survive text extraction - it is visible in the source PDFs linked below.

A formation sketch with a suspected course running between Budapest and Moscow is the single most intelligence-flavored item in the file, and it is also the item this page can say the least about, because the drawing only exists as an image in the PDF scans. What the text establishes is that the niece, or whoever drew it, believed the objects moved along a Budapest-Moscow axis - a detail that would have read very differently to a CIA analyst than "lights in the sky." Whether the sketch shows dots, discs, or arrows, readers can check against the original scans in about thirty seconds; the links are in the verification section.

Paragraph 3 adds the report's one piece of source-side calibration, normalized: "This is the first time my niece has mentioned such objects in her letters." The raw in the CIA-UAP-018 copy reads "~s :i,_s the first t:lme iq .niece has mentioned such obJeets in her J;etters." - the fragment "niece has mentioned such" is clean in that copy. It is a quiet sentence doing real work: the source is telling his CIA contact that this was not a correspondent who saw saucers everywhere.

How the CIA handled a family letter

The most revealing content in the file may be the handling instructions, because they show the machinery a piece of family mail entered once the CIA decided it was intelligence. Paragraph 4, normalized from OCR damage across both copies, reads:

"Available on loan from CIA Library is a photostatic copy of original letter in Hungarian and an English translation. The excerpt quoted in para. 1 is marked in brackets in this letter. UNCLASSIFIED"

Normalized reconstruction; the raw OCR is heavily damaged in both copies (CIA-UAP-013: "ffi.~le on loan f'rOlli CIA Li.brt:l.'ey is a ·i,b.otoatatic co_py at or1giml· letter..."; CIA-UAP-018 renders "Hungarian" as "lfuaorian" and "English translation" as "Ei~gl:!.'~h tr'.3.oP.l.a.t:1;.ott"). The phrases "on loan," "CIA," "quoted in," and "brackets in this le[tter]" are clean in at least one copy. Every repaired word is recovered from the surviving letterforms across the two copies; nothing is invented.

So the original Hungarian letter - the physical family mail - was photostatted and archived in the CIA Library, available on loan like a book, with the saucer passage bracketed for the next reader. The fuller English translation that appears as page 2 of the CIA-UAP-018 file matches the excerpt paragraph 1 flags, which strongly suggests, though the released pages do not label it, that this page is the very translation paragraph 4 describes. We flag that as our reading, not as a fact printed on the page.

Then there are the distribution controls. The report is marked (normalized from "NoFORN ,: NO DISSEM-ABROAD") "NOFORN: NO DISSEM ABROAD" - not releasable to foreign nationals, not to be disseminated outside the country. The fine print at the bottom of the form goes further, normalized from damaged but recoverable OCR in both copies:

"LIMITED: Dissemination limited to full-time employees of CIA, AEC and FBI; and, within State and Defense, to the intelligence components, other offices producing NIS elements, and higher echelons with their immediate supporting staffs. Not to be disseminated to consultants, external projects, or reserve personnel on short term active duty (excepting individuals who are normally full-time employees of CIA, AEC, FBI, State or Defense)..."

Normalized reconstruction from the CIA-UAP-013 extraction, page 1, cross-checked against the CIA-UAP-018 copy, which carries the same block with different OCR damage. The runs "employees," "and FBI," and "active duty" are clean in at least one copy. "Limited to" (raw: "li,njled to" in one copy, "limited fo" in the other) and "producing NIS elements" (raw: "producing NJS elemenf.r.." and "produC'ln9 NIS element,.") are repaired, as is the rest, from recoverable letterforms.

Read that plainly: a Budapest niece's secondhand account of saucer excitement was restricted to full-time employees of the CIA, the Atomic Energy Commission, the FBI, and the intelligence components of State and Defense. Not consultants. Not reservists. The AEC's presence is standard for the era's form language, but it underlines the frame: in 1955, unexplained fast-moving objects were handled inside the same channels as atomic-adjacent intelligence. And per the release stamp on the file, this stayed in restricted channels until it was, in the words of the CIA-UAP-018 header, "Approved for Release 2026 Under Section 1842 of the National Defense Authorization Act for Fiscal Year 2024" (raw, across two extraction lines: "Approved for Rel ease 2026 Und er Se ction 1842 of t he National Def ense" / "Authorizat i on Act for Fi scal Year 2024" - stray spaces break several words; the wording is otherwise verbatim). A less-redacted version of a 1955 family-letter report surfacing only in 2026 is its own quiet commentary on classification inertia.

CIA-UAP-009: the third Budapest file this site cannot quote yet

The release contains one more Hungary file: CIA-UAP-009, "Unknown Flying Objects Observed Over Budapest." The release metadata describes it as "a 1957 CIA Information Report regarding UFO sightings over Budapest" and notes, as with the other two, that a more redacted version has previously been available on CIA's public website. That is nearly everything this page can responsibly tell you about it, and here is why, stated plainly.

The automated OCR pass on the CIA-UAP-009 PDF produced an extraction file of just 140 bytes, page header included - seven short lines of punctuation-like marks with not one recoverable word. A representative line, in full: "-~---.....~...;~~..~ ...-•- ·i". That is the entire quality of the extraction. The PDF itself is small (27,826 bytes, displayed as 27.2 KB on its file page), and the scan is beyond what the current pipeline can read. This site's rule is that quotes come from text we can actually read, so until a manual transcription pass is done against the page images, we will not put words in this document's mouth. What can be said from the catalog data alone: it is a CIA Information Report, its event date is 1957 - two years after the letter, and months after Soviet forces put down the Hungarian Revolution in November 1956 - and it concerns objects over the same city. Whether it connects to the 1955 excitement the niece described is exactly the kind of claim we refuse to make without readable text.

What this is not

Honesty about the evidentiary chain here matters more than atmosphere. The 1955 report is hearsay at multiple removes: an unnamed estimator produced a speed figure; Budapest word of mouth carried it; the niece relayed it; her uncle freely translated it; a CIA officer transcribed the translation. Nobody in that chain is identified as a trained observer, no instrument recorded anything, and the report's own stamp says the CIA never evaluated it. The readable extractions contain no follow-up, no analyst comment, and no indication anyone at the Agency tried to identify what Hungarians were seeing in the autumn of 1955 - with the caveat that the CIA-UAP-018 file includes a third page whose extraction yielded no text at all, so this site cannot say what, if anything, is on it.

The document's value is therefore not "evidence of saucers." It is a primary-source record of three things: that a noticeable aerial-phenomenon excitement ran through Budapest for "several weeks" in late 1955, reported contemporaneously by a resident with no apparent stake in the matter; that the excitement was prominent enough for the letter to claim scientists were occupied with it; and that the CIA's collection system treated all of this as restricted foreign intelligence worth archiving. The "(rockets)" parenthetical is a standing reminder that mundane candidates existed even in the letter's own framing - and nothing in the readable extractions establishes or eliminates any of them.

Why 58/100

All three files carry an Anomalousness Index of 58/100 on this site's rubric, and it is worth being precise about what drives that number, because it is easy to misread. The strongest components are witness credibility, scored at the federal-agent tier (90, weighted 0.2) - which reflects the reporting channel, a CIA officer formally recording and disseminating the account, not the niece herself - and official disposition (90, weighted 0.1), because the record was released through a formal review process with no conventional-explanation finding attached. The weakest components are the ones the document candidly earns: sensor quality is eyewitness-only (30, weighted 0.25), since there is no instrument anywhere in this story, and kinematic anomaly scores as no-kinematic-data (30, weighted 0.15), because the lone speed figure is an unattributed popular estimate, not a measurement. Corroboration sits at the rubric's single_witness_instrument tier, meaning single-witness or single-instrument capture (60, weighted 0.2), and the mundane-explanation component is "weak mundane candidate" (70, weighted 0.1): rockets and misidentifications are conceivable, but nothing in the released record demonstrates one. Weighted and summed, that is 58. It is a statement about how much evidentiary weight this paper record carries - middling, and honestly earned - not a claim about what flew over Budapest.

How to verify everything on this page

1. Every quote above comes from this site's plain-text OCR extractions, which are linked from each file page. Where OCR damage was repaired, the repair is disclosed - in the note directly under the quote, or inline where the quote appears in a sentence - with the raw rendering shown; where damage was too severe to repair responsibly, we said so and did not quote.
2. The claim that CIA-UAP-013 and CIA-UAP-018 contain the same report is checkable in under a minute: open both extractions (or both PDFs) and compare the report number, subject line, source line, and the four numbered paragraphs.
3. The original war.gov PDFs: CIA-UAP-018, CIA-UAP-013, and CIA-UAP-009 (the "OBERVED" spelling is war.gov's own filename). The Budapest-Moscow sketch is visible on the scans.
4. File sizes and SHA-256 hashes for all three PDFs are published on the file pages linked above; they confirm the two copies of report 00-B-93674 are distinct files.
5. The 58/100 scores and component breakdowns are reproducible from this site's open rubric.