Bell 407 accident near Flagstaff, Arizona, February 5, 2026

The NTSB's narrative preliminary · quoted from the NTSB record

On February 4, 2026, at 2216 mountain standard time, a Bell 407 helicopter, N56AZ, was substantially damaged when it was involved in an accident near Flagstaff, Arizona. The pilot and tactical flight officer (TFO) were fatally injured. The helicopter was operated as a Title 14 Code of Federal Regulations Part 91 public use flight. The pilot and TFO began their 24-hour shift at 0800 on the day of the accident. They completed a previous flight to rescue hikers and returned to their base in Kingman about 2005. They subsequently received a request for a mission flight in Flagstaff involving an active shooter in a neighborhood southeast of downtown. The flight was performed with the TFO occupying the left-front seat and utilizing a forward looking infrared (FLIR) camera to assist law enforcement personnel in assessing the scene from the air. A review of Federal Aviation Administration (FAA) automatic dependent surveillance-broadcast (ADS-B) data, video and audio recorded by law enforcement body cameras, and residents’ security cameras revealed that the helicopter departed Kingman about 2110 and continued east toward Flagstaff. As the helicopter approached the city, the TFO was in communication with law enforcement personnel on the ground, who provided the approximate location of the active shooter. At 2205, the helicopter descended and operated between about 300 and 500 ft agl while performing a 360° turn over the neighborhood in the vicinity of the active shooter. The helicopter then performed a gradual climbing left turn south of the active shooter before establishing a northerly course. (See figure 1.) Figure 1: The helicopter’s flight path The pilot maneuvered the helicopter in a gradual climb north of the neighborhood and performed a left 180° turn to return south to allow clearance for a helicopter that was inbound to Flagstaff Medical Center. At 2212:10, while the helicopter was at an altitude of 9,950 ft msl, it passed in the vicinity of the hospital-bound helicopter, which was on a northeast route at about 7,200 ft msl. Thereafter, the helicopter performed a left 360° turn and returned south toward the law-enforcement scene while descending to 8,900 ft msl (see figure 2 below). The helicopter then turned north and performed a gradual climbing left turn, consistent with aligning for an out-of-ground-effect (OGE) hover to maintain visual contact of the suspect. At 2216:24, the TFO initiated a radio transmission to law enforcement personnel on the ground; there were no indications of any abnormalities during the communication. Figure 2: Flight path displaying airspeeds A review of available video showed the helicopter in a slow climb and then at 2216:43, the helicopter began a rapid rotation to the right. The rotation coincided with an unidentified verbal sound on the radio, followed by two loud banging sounds from the helicopter. The helicopter then descended rapidly toward terrain and at 2216:51, the TFO stated over the radio, “we’re going down.” During the descent, the helicopter’s spotlight was observed in video to flash periodically, consistent with continued aircraft rotation. The main wreckage was located on a flat dirt access road about 3 nm north of the Flagstaff Airport at an elevation of 7,010 ft msl. A section of the Burlington Northern Santa Fe (BNSF) railway tracks was located about 100 ft south of the main wreckage. The wreckage was distributed along a debris path on a median bearing of 250°; the most distant wreckage was located more than 6,150 ft from the main wreckage. The main wreckage consisted of the cockpit, cabin, and aft fuselage, which came to rest on its right side. The cockpit and cabin exhibited extensive postaccident fire damage. The tail rotor, vertical stabilizer, and a section of the tailboom came to rest about 135 ft northwest from the main wreckage. The tail rotor driveshaft remained attached to the tail rotor gearbox and exhibited angular cuts at the forward section; the surrounding skin exhibited deep grooves with blue paint transfer consistent with contact with a main rotor blade. The tailboom was fractured at its forward location just aft of the intercostal support, with the structural remnants of the upper-right attachment fitting exhibiting bending outboard. A section of forward tailboom and connected aft fuselage were torn away from the fuselage with one tailboom attach bolt intact connecting the two halves. The remaining three tailboom attach bolts were intact at the tailboom/fuselage joint. The tailboom and the upper portion of the vertical fin exhibited multiple impact marks and punctures consistent with main rotor blade strikes sustained during the separation sequence. A comprehensive review of the aircraft's logbooks and maintenance records established that there had been no recent maintenance actions, reported anomalies, or structural repairs involving the tail boom or tail rotor assemblies prior to the accident sequence. The main rotor hub, main rotor transmission, and three main rotor blades were located about 165 ft south of the main wreckage (see figure 3 below). The fourth main rotor blade, blade grip, and portion of its respective yoke were not located with the main rotor assembly; it was separated from the rotor hub and located 650 ft northeast of the main wreckage. All blades were lightly coated with an orange/yellow oil film. Postaccident examination of the helicopter revealed no evidence of ballistic punctures. A review of video and ADS-B data revealed that the helicopter’s closest proximity to the suspect at the time shots were fired was about 7,500 ft. Figure 3: Picture of wreckage distribution and composite of video The engine control unit (ECU) was recovered from the wreckage, and its data was successfully downloaded (see figure 4 below). A review of the recovered log revealed 16 trend records and 4 high-resolution snapshots, totaling 20 lines of data over an 18.4-second period. For the initial 6 seconds of the recording, the helicopter was operating at high power; the gas generator speed (Ng) was between 100% and 101%, and the main rotor speed (Nr) remained stable at 100%. Engine torque (Q) was between 76% and 80%, and the collective position (CP) was 64%. The power lever angle (PLA) was constant at 66°, which was consistent with the FLY detent range of 62° to 70°. During the following 1.2 seconds, Ng remained at 101%, Nr increased to 101%, and fuel flow (FF) decreased to 296 pph. During the next 1.2 seconds, the collective was reduced to 48%, and fuel flow decreased to 284 pph. Simultaneously, Ng reduced to 97%, and Nr and power turbine speed (Np) both reduced to 98%. The following 1.2-second interval, beginning 9.6 seconds after the initial record, showed Nr and Np at 98%, while Ng increased to 102% and torque increased to 92%, with a fuel flow of 420 pph and an increase of measured gas temperature (MGT) to 1460°F. The next recorded data, 1.2 seconds later, showed that Ng decreased to 95%, Nr to 97%, Np to 96%, and torque to 60%; the collective was reduced to 10%. During the subsequent 1.1 seconds, spanning four lines of data, the ECU recorded a transition to the reversionary governor. Nr decreased to 0%, while Np increased to 112% (overspeed). Simultaneously, engine torque decreased to 2%, consistent with an instantaneous removal of the rotor load from the power turbine. During the final 5.2 seconds, the PLA was reduced to 28°. The engine core remained operational with Ng stabilized at 63%, which is the specified ground idle speed. Figure 4: Excerpt of ECU downloaded data According to the operator, the pilot would have been operating with night vision goggles. The pilot’s total flight hours are unknown, although he reported to the FAA that he had amassed more than 3,500 hours flying civil helicopters. He served in the Marine Corps from 1988-2010, a portion of his service was attached to Marine Helicopter Squadron One (HMX-1) as a pilot. The pilot of another helicopter, which was en route to land at the hospital at the same time the accident helicopter was operating in the area, stated that at 7,200-7,500 ft msl the wind was from the northeast at about 10 to 15 kts. A high-resolution rapid refresh (HRRR) model over the time of the accident displayed that the wind between about 9,500-10,200 ft msl was about 20 kts from 095°-115°. According to the U.S. Naval Observatory, Department of Astronomical Applications, the phase of the moon was waning gibbous (just past full) with 87.6% of the moons visible disk illuminated. Moonrise was at 2113 and at the time of the accident the moon was about 11° above the horizon (including refraction) on an azimuth (direction) of 99°.

The complete narrative as the NTSB published it. The NTSB's docket holds the report as a PDF and any photographs, statements and other documents from the investigation.

About this page