GEIPAN’s dynamic statistics, current as of 28 April 2026, cover 3,351 classified and published cases. The headline number is 3.2% — 106 cases that remain genuinely unexplained after full investigation. Here is the full breakdown and what each category means.
The Classification Breakdown
- Category A — 27.81% (932 cases): Phenomenon perfectly identified. Full explanation established beyond ambiguity.
- Category B — 38.76% (1,299 cases): Phenomenon probably identified. The GEIPAN hypothesis is considered very probable but not definitively confirmed.
- Category C — 30.26% (1,014 cases): Phenomenon unidentified due to lack of data. Not enough information was available to reach a conclusion. This is not the same as “unexplained” — it means the evidence base was insufficient.
- Category D — 3.16% (106 cases): Phenomenon unexplained after full investigation. GEIPAN conducted an investigation and could not identify a known explanation.
What Category D Actually Means
Category D is further divided. D1 cases are unexplained but have “medium consistency” — typically a single witness, no photographic or video documentation. D2 cases have “strong consistency” — multiple independent witnesses and/or photographic or video recording and/or physical ground traces. D2 is the highest evidential standard in GEIPAN’s system. The dynamic statistics list D2 at 0.00% in the current published set, meaning no currently published cases meet the highest evidential threshold for unexplained phenomena — though this can change as new cases are classified.
The 106 Category D cases across 49 years are not reports that were dismissed. They received field investigation — meaning a GEIPAN investigator met the witness, reviewed the evidence, consulted external experts, and still could not identify a known explanation. Only about 10% of all GEIPAN cases receive field investigation. The D cases are the residue of that rigorous process.
What Explains the Explained 66.5%
Of Cases A and B combined (66.57%), the primary explanatory families documented in GEIPAN’s 2023–2025 sample are: aeronautical phenomena (conventional aircraft, contrails, helicopters, airships — excluding drones), balloons (weather balloons, stratospheric balloons, LED balloons, Mylar), astronomical objects, atmospheric phenomena, and drones. The balloon category in particular has grown significantly with the proliferation of consumer and commercial aerial balloons and Starlink-related observations — GEIPAN issued an explanatory note in July 2024 specifically on Starlink satellite flaring, which had been generating large numbers of reports from commercial pilots.
The Oracle Assessment
106 cases unexplained after full investigation in 49 years of rigorous French state analysis. This is not a small number dressed up as significant. In absolute terms it is modest. In scientific terms it is a residual that cannot be explained away by misidentification, sensor error, or witness unreliability — because those explanations were specifically tested and found insufficient for these cases. The French government has been sitting on 106 genuinely unexplained UAP cases, published and publicly accessible, since at least 2007. The global UAP conversation is only now catching up to what GEIPAN documented decades ago.
Source: cnes-geipan.fr/fr/stats. GEIPAN dynamic statistics, 28/04/2026. Classification page, 31/03/2026.