An Air Force RB-47H crew reported unusual lights and electronic signals during a flight across the southern United States on 17 July 1957. The best-known part of the encounter developed around Dallas–Fort Worth and Mineral Wells, Texas. Its significance lies in the combination of visual observations, intercepted radio-frequency signals and reported ground-radar contact. Its difficulty lies in deciding how firmly those different observations were connected, and in separating the 1957 documents from recollections collected a decade later.
The date is July, not September
The surviving Air Defense Command letter dated 15 August 1957 asks Major Lewis D. Chase to complete an observer’s data sheet concerning his sighting over Fort Worth on 17 July. That is direct contemporary support for the incident date. The wing’s message also gives 17 July and identifies the aircraft by its call sign, Lacy 17. These records are a better date anchor than a later attempt to reconstruct the season from memory.
The Colorado study’s Case 5 was investigated through interviews in 1967 and appeared as an autumn 1957 event. Its investigator, Roy Craig, could not locate the original report. James McDonald’s later published account explains that the file search and weather work had used an incorrect September date. Locating the July file therefore changed the evidential situation: the case no longer depended entirely on ten-year-old memories. It did not, by itself, establish what caused the observations.
A light and a signal were different kinds of observation
The 55th Strategic Reconnaissance Wing’s four-page message describes a planned track from Meridian, Mississippi, towards Waco, Texas. An electronic operator intercepted a signal near 3,000 megacycles per second, equivalent to about 3,000 MHz. The message gives its pulse characteristics and describes movement across the direction-finding display. It interprets that movement as indicating an airborne source. That is the report’s interpretation of a received signal, rather than a photograph of an aircraft.
The passive equipment listened to incoming transmissions and gave their direction. It did not work in the same way as a radar transmitting a pulse and detecting its reflection. Calling every electronic observation an independent radar lock would conceal precisely the question under investigation: whether the signals came from an unfamiliar airborne source or from ordinary ground radar received under confusing circumstances.
At 1010Z, the message records a very intense white light with a pale blue tint crossing in front of the aircraft from the commander’s left towards his right. The co-pilot also saw it. Later signal observations are timed separately, including bearings at 1030Z, 1035Z and 1038Z. This sequence matters when assessing claims that every visual and electronic event was exactly simultaneous.
The Texas sequence in the contemporary message
At 1039Z the commander reported a large light apparently below the RB-47, which was at 34,500 feet. The message explicitly says he could not determine the object’s shape or size. His impression that light came from its upper part should not be converted into a measured solid structure or a particular craft design.
At 1040Z the message records two electronic bearings and says the commander and co-pilot saw two red objects at the same time. Permission was obtained to depart from the flight plan and pursue the apparent source. Subsequent entries alternate between one and two signals. These are specific reported details, rather than a reason to imagine a whole fleet or assign a geometric formation.
Around 1048Z the crew reported an object about ten nautical miles northwest of Fort Worth, and the message says the ground station, code-named Utah, confirmed it on its scopes. Around 1050Z the object appeared to stop, the aircraft overshot, and both Utah and the electronic operator lost contact. During a turn, the report says visual contact and the other indications returned.
The account then describes a further loss of contact as the apparent object dropped lower. At 1055Z, in the Mineral Wells area, the crew advised Utah that fuel required a return home. At 1057Z an electronic signal was still being received while Utah had no scope contact. A later visual report and further signals followed before the northbound departure. The record thus includes episodes of agreement and episodes in which the different channels did not all show the same thing.
An Air Force analyst’s caution in October 1957
A 30 October 1957 internal analysis, signed for Captain Edwin Mammen, describes the report as difficult to evaluate. It acknowledges the apparent connection between the observations while emphasising that the intercepted signals had the characteristics of ground radar, resembling the CPS-6B type. The office knew of no airborne S-band equipment with the stated characteristics.
The same analysis says that the type of equipment at Utah was not known to the writer and that there was no firm correlation between the ground intercept and the aircraft sightings. It could not make a determination from the information submitted. This contemporary caution is more informative than treating either the unfamiliar signal or the officer’s rank as sufficient proof of an extraordinary vehicle.
The wing’s message had expressed confidence that electronic bearings and visual observations coincided. The later technical note was less certain about correlation. Both positions survive in the file. Their disagreement should remain visible rather than being flattened into a single official verdict.
What later interviews added—and could not settle
Craig interviewed the pilot, co-pilot and the most involved monitoring officer. The navigator was unavailable, and two other crew members were not located. His report records substantial disagreement about whether film or recording wire had even been aboard and about post-flight questioning. Its conclusion was that recollections alone could not identify the phenomenon. Those limitations remain relevant even though the original written reports were subsequently located.
McDonald’s 1971 account drew on the recovered file and his own interviews. He regarded the combination of visual, electronic and ground-radar reports as particularly compelling. He also identified differences: some details of the early report were not recalled in the interviews, including an occasional second source. His interpretation is an investigator’s argument about the evidence, not an independent repetition of the flight under controlled conditions.
The airliner explanation needs two distinctions
The compilation includes a Civil Aeronautics Administration letter dated 8 November 1957 about two commercial-airliner incidents. One was near Amarillo on 22 July; the other involved American Airlines Flight 655 near El Paso on 17 July, with Flight 966 identified as the other aircraft. The letter does not describe Chase’s RB-47 flight near Fort Worth. Its presence in a case compilation does not make its conclusion automatically applicable to every episode in that compilation.
Philip J. Klass’s separate analysis of December 1971 proposed a combination of conventional causes: ground-radar transmissions for the electronic signals, a meteor for the first fast light, and ordinary lights or air traffic for later observations. He suggested Flight 966 as a possible Dallas-area radar target, while acknowledging that he could not obtain the original flight records to verify its timing.
Klass also reproduced favourable comments from Chase and electronic operator Frank McClure about his explanation. These are reported exchanges in Klass’s paper; the original correspondence was not independently examined here. They complicate a simple story of witnesses unanimously rejecting conventional explanations, but do not supply missing recordings or establish every proposed identification.
A documented report with unresolved associations
The case has an actual 1957 paper trail, and reported radar involvement belongs in its description. What remains uncertain is the identification and association of the individual lights, received signals and ground returns. The original records are stronger evidence of what was reported than later summaries, yet even they contain interpretive judgements and gaps. This article does not claim a fresh reconstruction of antenna coverage, flight tracks or raw radar data.
The map pin is an approximate Fort Worth city reference from the Census gazetteer. It represents the Texas portion of a much longer flight through several southern states, not an exact airborne position, a signal transmitter or the aircraft’s Kansas home base. The incident year is 1957; the dates of later studies are not additional encounters.
Sources and further reading
Compiled RB-47 file, via The Black Vault: original Air Force letter at PDF page 3; wing message at pages 37–40; technical analysis at page 55; CAA letter at page 57. These original documents were read separately from the compilation’s copyrighted later material. The compilation also includes Philip J. Klass’s The RB-47 UFO Case—A New Explanation, dated 30 December 1971, of which the relevant discussion and conclusions were reviewed.
Colorado study, Case 5, investigator Roy Craig; James McDonald’s 1971 Astronautics & Aeronautics account, reproduced by NICAP; Census 2025 national places gazetteer. Sources checked 10 November 2026.