THE DOCUMENT
DOW-UAP-D099, “Intelligence Review of Ghost Rocket Incidents, 1947”, a Department of War record from the U.S. Department of War’s PURSUE Release 05, published at war.gov/ufo, cleared 7 August 2026. Eight pages, and an extract rather than an issue. The cover reads INTELLIGENCE REVIEW, Number 47, 9 Jan 1947, issued by the INTELLIGENCE DIVISION, WDGS, WAR DEPARTMENT, Washington, D. C., Copy No. 597. The masthead, the issue number and the date did not survive OCR and are readable only on the page image; what the released text layer carries from the cover is “Copy No. 597” and the WDGS line. SECRET is printed at the top and bottom of the cover, alongside a red boxed receipt stamp dated 2 0 JAN 1947, and struck through by hand on the contents page and on all four article pages. The declassification stamp reads DECLASSIFIED, Authority NND58658, By BKT, NARA, Date 6/16/26. The eight pages are the front cover; a contents page listing nine items running to printed page 64; the feature article on printed pages 17 to 20; the Notice and DISTRIBUTION page; and a back cover blank but for one struck-through SECRET. The rest of the issue is not in the release.
Why this one is worth your time
Every other record in Release 05 is dated later than this one. The Air Materiel Command file in Release 05 Briefing 5 opens on 23 September 1947, the Navy’s film analysis is 1953, the State, CIA and Executive Office papers run 1963 to 1965, and the FBI interviews and video records run 2002 to 2026. D099 is dated 9 January 1947, the earliest-dated record in the tranche by more than eight months.
It is also a different kind of paper from nearly everything around it: not a sighting report, not an interview, not a working file, but a finished intelligence product. It is a classified weekly digest produced by the Intelligence Division of the War Department General Staff, printed, numbered by copy and circulated from the White House to the Marine Corps. The ghost rockets sit in it as one feature item among nine, between “WORLD IN REVIEW” and a piece on the British Territorial Army. Four printed pages later it has reached its own conclusion: two or three real incidents, perhaps as many as five or ten, with meteors or fireworks for the rest.
What the document says
Where the article sits. The contents page lists it second, “GHOST ROCKETS” OVER SCANDINAVIA at page 17. The article opens under a boxed FEATURE SECTION rule, carries the running head GHOST ROCKETS on the pages after the first, and runs under four subheads, the last Conclusions. No author is named.
The wave, and the descriptions. The press first reported flying missiles over southern Sweden in late May 1946 and supplied the name; by July Swedish sightings had greatly increased, with several more from Norway. The great majority came from untrained observers and, the article says, varied widely. The two commonest descriptions were “a ball of fire with a tail” and a “shiny cigar-shaped object”. Reported direction covered all points of the compass, northerly slightly predominant; altitudes ran from treetop height to 160,000 feet, the higher figures almost exclusively from Finland; speeds from 65 m. p. h. to “lightning fast”. A few objects were reported diving into the ground or into lakes, or exploding in the air.
The one witness the article rates, and then discounts. One report is singled out as from someone who should be more reliable and competent than the average layman: a Swedish Air Force pilot at 650 feet over central Sweden at 1000 hours on 14 August, who saw a dark cigar-shaped object about 50 feet long and 3 feet in diameter, roughly 6,500 feet away, at an estimated 400 m. p. h., with no visible wings, rudder or projecting part and none of the flame or light reported in most other sightings. The article then takes most of it back: the pilot said the missile held a constant altitude over the ground and consequently followed the terrain, which the article says casts doubt on the reliability of the entire report, a wingless missile being unable to hold constant height over hilly ground. It allows that he implies wings he could not see, having ruled out a Swedish jet because none was flying nearby.
The press. The Scandinavian press, the Communist papers excepted, initially reported the incidents in detail and openly attributed them to missiles fired by the U. S. S. R. A partial censorship in August restricted publication of exact details and of the localities where missiles were seen. The Communist press ridiculed the matter, denied any Soviet basis for it, and charged that the missiles came from the United States and that Gen. Doolittle had been sent to observe their effects.
The investigation, the communiqués and one footnote. Official investigations began in June, when the Defence Staff asked the public to report any unusual observations. The article calls the investigation very peculiar: key personnel were on summer leave at the start and were not recalled. Spokesmen repeatedly told the United States Military Attaché that Soviet rockets launched from Peenemünde were over Sweden, and were never able to produce evidence for it; nor had any United States military or naval personnel in Sweden seen fragments, points of impact or other direct evidence. The Defence Staff’s public communiqués matched neither those statements nor each other: the one of 6 August reported that except in a few cases the objects could not be meteorites; the last, of 10 October, said the opposite. The article reads the reversal as the Swedish Staff minimising a matter grown large enough for it to fear official repercussions on Swedish-Soviet relations, and says no definite conclusion can be drawn from the conflicting evidence. A footnote to that passage is the only place the article leaves the north: in September and October flying objects were reported over widely separate points in Europe and Africa, including Belgium, Greece, Italy, Morocco and Austria, in the main unconfirmed and put down to explainable causes such as meteors, Very flares and fireworks.
Where it thinks they came from, and how it ends. The Soviets were known to be working on guided missiles and had probably tested the V-1 type; without warheads and with slightly improved motors those could range 500 miles, and were the only available German missiles believed capable of horizontal flight at low altitudes. There was some evidence of such work at Stolp, in Polish-administered Germany, or on the Baltic islands of Oesel or Dago. Three of those reporting sightings mentioned the noise of an outboard motor, a characteristic of the impulse duct motor the Germans used on the V-1, and one said what he saw looked like the V-1 he had seen over London during the war; others reported little or no sound, which may indicate a turbo-jet instead of a pulse-jet. The high-altitude reports seem definitely to have been meteors or fireworks, and the Swedish Defence Staff probably took advantage of the situation for political purposes, letting the newspapers build the story at a time when the Swedish public was demanding cuts in defence expenditure.
The back matter. The last printed page carries a Notice asking any reader with information that appears to modify or contradict the intelligence to forward it to the Director of Intelligence, WDGS, War Department. Below that is the DISTRIBUTION table: 141 copies to the Army Air Forces, 58 to military attachés, 16 to the State Department, 12 to the Central Intelligence Group, 3 to the FBI and 2 to the White House.
What the document does not say
There is no radar in it. The word does not appear once in the eight pages. Every observation it handles is somebody looking at the sky, and its reliability language is about eyesight and competence, not instruments.
Almost nobody is named. The pilot is a Swedish Air Force pilot, the spokesmen are spokesmen, the Military Attaché is a title. The one personal name is Gen. Doolittle, inside a claim the article attributes to the Communist press.
The nonmetallic slag arrives with nothing attached: who found it, where, when, who examined it and on what basis it was called nonmetallic are not in the article.
The report does not source itself. No document is cited, no attaché report numbered, no reference given, no analyst credited. The only statement about sourcing in the extract is the Notice’s, that the intelligence was produced by the Intelligence Division from careful research and analysis of information received from all sources.
Only one date attaches to any individual sighting, the pilot’s 14 August. Nothing dates the almost 1,000 public reports beyond the end of July, nothing records what the Defence Staff did with them, and nothing connects any of them to the 2 or 3 the article ends on, a figure offered as a judgement rather than a list.
And the rest of the issue is not here. Eight of the nine items on the contents page are absent, among them “GUIDED MISSILE COUNTERMEASURES” at page 50 and “MILITARY IMPLICATIONS OF SOVIET FOREIGN POLICY DURING 1946” at page 41.
From the record
Flying missiles were first reported over southern Sweden in
DOW-UAP-D099, printed page 17, the opening sentences of the article; the paragraph continues that by July the number of sightings over Sweden had greatly increased, and several also had been reported over Norway
late May 1946 by the press, which gave the missiles the name
of “Ghost Rockets.” In June, the missiles also had been
reported over Finland and Denmark.
In no case have fragments been found other than bits of
DOW-UAP-D099, printed page 17, closing the paragraph on how the objects were described; the article does not come back to physical evidence
material described as “nonmetallic slag.”
end of July almost 1,000 such reports had been received.
DOW-UAP-D099, printed page 19, on the returns from the Swedish Defence Staff’s appeal to the public; the sentence begins at the foot of page 18: the Defense Staff requested the public to report any unusual observations, and by the
for the Defense Staff repeatedly have told the United States
DOW-UAP-D099, printed page 19, the attribution Swedish officers gave the U.S. Military Attaché in conversation; the next sentence records that they had not been able to produce any evidence to support those statements. The page prints Peenemünde with an umlaut and Attaché with an accent; the released text layer preserves neither
Military Attaché that they definitely believed there were
rockets over Sweden, and that they were launched by the
Soviets from Peenemünde on the German Baltic coast.
issued on 10 October, very strongly indicated that the great
DOW-UAP-D099, printed page 19, the last of the Defence Staff communiqués; the sentence continues that this is a complete change from the one of 6 August, which had reported that except in a few cases the objects could not be meteorites
majority of the reported incidents were of celestial origin,
The best evidence, at present, is that there have been only
DOW-UAP-D099, printed page 20, opening the Conclusions section; the sentences that follow say the high-altitude missiles reported seem definitely to have been meteors or fireworks
2 or 3 real incidents, perhaps as many as 5 or 10, of low-flying
missiles of the V-1 type.
Where the case connects
Release 05 holds this record and, eleven months later, the Air Force’s decision that it was a separate matter. The Air Materiel Command file in Release 05 Briefing 5 contains an “Analysis of ‘Flying Disc’ Reports” of 18 December 1947 whose first paragraph states that no connection was found between the Swedish ghost rocket sightings and those in the United States and the Pacific. It also names a “Swedish Incidents” file, requested in January 1948 and never answered in its 245 pages.
That distribution put the article in front of readers who turn up elsewhere in PURSUE. The Central Intelligence Group was the body the CIA succeeded later that year, and Release 03 covers the CIA historical block. Nothing in this release records what any of those addressees did with the article.
Release 05 Briefing 4 covers the Navy’s 1953 film analysis, six years on and a different service, which measures two films instead of assessing reports and likewise ends with no physical evidence. D099 works the other way round, on nearly a thousand accounts and no instrument data at all, and reaches its figure by subtraction. Its loose ends, the slag, the almost 1,000 reports and the attaché reporting behind all of it, are named in the article and are not in this release. Any later tranche that carries them lands in this series.
Read it yourself
DOW-UAP-D099, “Intelligence Review of Ghost Rocket Incidents, 1947”, eight pages, is hosted at war.gov in PURSUE Release 05. The article is PDF pages 3 to 6, printed pages 17 to 20. Open the cover as an image, not as text: the masthead, issue number and date are on the page and not in the text layer.
Read the file. Decide for yourself.
Related wiki entries
The wiki entries below give background on the programme and publisher behind this briefing, and on the subjects it touches.
References and further reading
- DOW-UAP-D099, “Intelligence Review of Ghost Rocket Incidents, 1947”, PURSUE Release 05, U.S. Department of War, hosted at war.gov/ufo
- Presidential Unsealing and Reporting System for UAP Encounters (PURSUE), U.S. Department of War, war.gov/ufo
- AARO UAP Records, All-domain Anomaly Resolution Office, aaro.mil/UAP-Records
- Signals from the Periphery, Release 05 Briefing 5, on the Air Materiel Command file and its December 1947 finding on the Swedish sightings
- Signals from the Periphery, Release 05 Briefing 4, on the Navy’s 1953 film analysis
- Signals from the Periphery, Release 03, on the CIA historical block