On October 2, 2003, approximately 0740 Universal Time Coordinated (UTC), a Boeing 747-100, N479EV experienced an in-flight fire in the fire detection card file while in cruise flight approximately 35 miles from Lyon, France. None of the three crew members on board were injured, and the damage to the aircraft was limited to the heat/fire damage within the card file container. The aircraft, which is owned and operated by Evergreen International Airlines, was in visual meteorological conditions at the time of the incident. The aircraft was on an IFR flight plan from Kuwait City, Kuwait, to Frankfurt, Germany. The aircraft departed Kuwait City at 0148 UTC. According to Evergreen Airlines, while the aircraft was in cruise flight at flight level 400, the right wing overheat light illuminated. The crew accomplished the appropriate checklist, and then about five minutes later the engine number three and engine number four B Loop fire indicators illuminated. Soon thereafter an acrid odor was detected by the flight crew, and smoke was seen coming from behind the P-6 circuit breaker panel. The crew then became aware of a small fire in the fire detection card file, which they extinguished with a hand-held fire extinguisher. The crew then diverted to Lyon, France, where an uneventful normal landing was executed. After landing at Lyon, French civil aviation personnel came aboard the aircraft, interviewed the crew, and then departed. Prior to the aircraft's departure from Lyon, the entire card file was removed and an airworthy unit was installed, and circuit breaker C431, which protects the A16 card, was replaced. The fire detection card file was then returned to the United States for analysis. After its arrival in the United States, the card file was eventually shipped to Boeing Equipment Quality Analysis (EQA), where a series of inspection and test activities were performed under the guidance of the NTSB. The initial inspection of the fire detection card file (Boeing P/N 65B47509-33) revealed severe thermal stress/destruction to card A16 (engine number four B Loop fire detection), and minor thermal discoloration to components adjacent to this card. Card A16 was manufactured by Walter Kidde (S/N 4538) and is identified as Boeing P/N 60B00023-96. The thermal distress on card A16 included melted/burned circuit traces, failed IO (transient protection) devices, partial or total destruction of a number of the card-to-file connecting pins, and burning of the phenolic board itself (see photo #1). In addition there was severe dishing under portions of the K2 relay, an orange residue in the area of the destroyed card pins, and two globs of unidentified material near the point where one of the destroyed connecting pins attached to its circuit trace near the edge of the card (see photo #2). The area of maximum thermal distress on the card was located in an area between the inboard half of the K2 relay and the inboard ends of the third and fourth card-to-file connecting pins at the K2 end of the card. In addition to the damage to card A16 itself, there was also severe thermal damage to the card file electrical harness socket connector that the A16 card was connected to. This damage included melting/burning of the connector structure, loose sockets still connected to their associated wires, missing sockets, and open wires that had been attached to sockets that were now missing (see photo #3). Further investigation determined that the proper fire detection element was installed on engine number four of the subject aircraft, and that the C431 circuit breaker installed in the aircraft at the time of the event was of the correct 2.5 amp rating. In addition it was determined that the two globs found on the card, as well as the orange residue, where consistent with material used to manufacture the card, and not consistent with evidence indicating foreign materials. It was also determined that there was no clear evidence of an initiating short in the aircraft's systems, the card file, or the card itself. The investigation was also not able to find any indication that the aircraft had been struck by lightning during this flight, or that a lightning strike had contributed to the initiation of the card fire. Furthermore, regardless of the fact that card A16 displayed indications of extreme heat distress, there was no clear evidence indicating that an anomaly or malfunction in any of the sub-components contained on the card itself initiated the sequence that lead to the overheat event. After a series of follow-up conference calls and additional tests, the investigative team determined that they were not able to identify a specific component or event that initiated the sequence that lead to the card fire. According to Evergreen Airlines, the subject aircraft has continued in service since the time of this event, and has generated no other maintenance discrepancies known to be directly related to the fire detection card files.