Cussac UFO encounter

Case dossier

The firm chronology begins on Tuesday, 29 August 1967. François and Anne-Marie were tending the family’s cattle at Les Tuiles, west of Cussac, when their dog reacted as a cow approached the field boundary. At roughly 10:30, François noticed small dark figures beyond a hedge and drew his sister’s attention to them. Their father, who was mayor of Cussac, contacted the gendarmerie after the frightened children returned. These are reported actions preserved by interviews and early press, not independently timed sensor events.

The earliest public version appeared in La Montagne on 1 September. It described four ‘dwarfs’ in black coveralls, facial hair, a dazzling object estimated at about four metres by two, the figures diving headfirst into it and one apparently coming back out before departure. Crucially, the article also said nearby people working in fields heard nothing and no trace had been found. Later accounts changed the geometry toward a sphere, added more precise motions and enlarged the physical-effects narrative. The early article is therefore an anchor, not a complete transcript of separate witness interviews.

Local gendarmes did visit, but the public provenance is incomplete. The original 1967 procès-verbal has not been located in the reviewed online record. The 1978 GEPAN report says a re-interviewed gendarme remembered a strongly sulphurous odour and a four-to-five-metre area of slightly yellowed grass. An earlier civilian GEPA investigation also attributed a sulphur smell to police. Because no contemporaneous sampling, photograph or laboratory result survives publicly, the smell and vegetation claims support a history of reported effects, not a verified landing trace.

The civilian GEPA inquiry published in 1968 and other specialist retellings fixed many details before the official CNES group existed. GEPAN was created only in 1977. Its four-person team visited Cussac in April 1978, more than ten years after the event, interviewed the siblings separately, used a theodolite and timed reconstructions, and spoke with indirect witnesses. One rural guard then recalled a whistle or droning sound from several hundred metres away. That late recollection conflicts with the earliest press statement that nearby adults heard nothing and cannot be treated as contemporary corroboration.

GEPAN’s internal assessment was favorable to the witnesses: the investigators considered them sincere, found the reported phenomenon unlike known categories available to them and regarded a piloted vehicle as fitting the collected description. That conclusion must be read beside GEPAN’s own public methodology brochure, which said its 1978 investigations of old cases allowed only limited conclusions because physical measurement and precise psychological or social reconstruction were no longer possible. Institutional attention and a strong investigator interpretation do not convert the report into confirmation of a craft or occupants.

The 1999 COMETA report later summarized Cussac under the heading ‘GEPAN/SEPRA inquiry’ and repeated the favorable reconstruction. COMETA was a private committee report hosted in GEIPAN’s document library, not a contemporaneous 1967 finding and not a new SEPRA field investigation. Its wording has often been repeated as ‘official confirmation,’ obscuring that the event account, 1978 review and 1999 advocacy document are separated by decades and have different evidential roles.

Conventional explanations do not produce a clean identification either. A helicopter with crew can resemble a rounded bright body and dark figures, and reported military activity made that hypothesis worth testing; however, no specific aircraft, crew, route or flight log has been tied to the field. Memory consolidation, interviewer influence and a brief unfamiliar stimulus can explain changing detail without accusing the children of fabrication. The responsible conclusion is narrower: two children made an early, striking report; later investigators added support and interpretation, but the surviving record cannot determine what physical stimulus, if any, generated it.