Introduction
Comair Flight 5191 crashed shortly after 6:06 a.m. on August 27, 2006, after attempting to depart from the wrong runway at Blue Grass Airport in Lexington, Kentucky. The Bombardier CRJ-100 had been cleared for Runway 22, the airport’s 7,003-foot primary runway, but the crew instead entered Runway 26, an unlighted 3,501-foot general-aviation runway. The aircraft accelerated, overran the pavement, became briefly airborne, struck terrain and trees, and was destroyed by impact forces and fire. Forty-nine of the fifty people on board died; the first officer survived with critical injuries. The National Transportation Safety Board concluded that the crew failed to use available cues to identify the airplane’s position during taxi and failed to cross-check and verify the runway before takeoff (NTSB, 2007). The accident is especially instructive because no single mechanical defect explains it. It emerged from an interaction among crew attention, expectation, cockpit discipline, airport construction, visual cues, controller monitoring, and missing technological defenses.
Wrong-Runway Entry and the Missed Cues
The flight departed before sunrise while Blue Grass Airport was undergoing construction that had altered parts of the taxiway environment. The crew had been cleared to taxi to Runway 22 and possessed several independent cues that could confirm the intended departure surface: airport diagrams, signs, markings, runway lighting, the clearance itself, and heading information. During taxi, however, the airplane entered Runway 26 instead. Once aligned, the absence of expected runway lighting became an important contradictory cue, and cockpit voice recorder evidence showed that the first officer commented on the darkness. That observation did not lead to a full stop and re-verification. This is central to the accident because safety systems rely not only on the presence of information but on procedures that force crews to act when information conflicts with expectation. The crew’s mental model had become strong enough that contradictory evidence was interpreted or tolerated rather than treated as a reason to abandon the takeoff. A conservative verification step could have broken the sequence before power application.
Expectation, Confirmation Bias, and Crew Resource Management
Human performance research helps explain why trained pilots can share the same mistaken belief. People interpret ambiguous environments through expectations, and once a crew believes it has reached the assigned runway, cues may be unconsciously fitted to that assumption. Confirmation bias can make supporting information feel more important than contradictory evidence. In Flight 5191, the intended runway had already been mentally established, and both pilots proceeded as though the expected location had been reached. Crew resource management is designed to counter exactly this problem by converting assumptions into explicit cross-checks. Its purpose is not merely for one pilot to correct another after an obvious mistake; it is to create structured communication that can reveal a shared error. A robust departure procedure requires both pilots to identify the runway designation and compare the actual heading with the expected heading before takeoff power is applied. If either pilot is uncertain, the safe response is to stop, clarify position with air traffic control, and accept delay rather than continue with unresolved runway identity.
Sterile Cockpit, Attention, and Fatigue
The NTSB identified nonpertinent cockpit conversation during taxi as a contributing factor. Federal sterile-cockpit requirements restrict nonessential duties and conversation during critical phases of flight because attention is limited and even ordinary discussion can interrupt monitoring, checklist completion, or interpretation of signs. In this case, construction had made the surface environment less familiar, increasing the importance of protected attention. Fatigue was also examined because the flight operated during an early-morning period when alertness can be reduced, yet the NTSB did not establish fatigue as the probable cause. That distinction matters. Accident analysis should separate a plausible performance influence from a factor supported strongly enough to appear in the formal finding. The broader lesson is still relevant: airlines should manage circadian disruption, rest opportunities, commuting, schedule changes, and cumulative workload rather than treating alertness solely as an individual responsibility. Sterile-cockpit discipline and fatigue-risk management are organizational defenses that make human performance more reliable when environmental complexity is high.
Airport, Controller, and Technological Defenses
The controller issued the correct clearance and did not instruct Flight 5191 to use Runway 26, yet system safety asks whether independent defenses could have detected the error before takeoff. The controller was performing other duties after issuing clearance, and the airport did not have surface-surveillance equipment capable of automatically warning that the aircraft had entered the wrong runway. Construction also changed visual patterns and taxi expectations, creating a temporary environment that required especially clear markings, diagrams, notices, and nighttime evaluation. Modern safety strategy uses defense in depth because crews, controllers, and technology can each fail. Airport moving maps can display own-ship position, runway-awareness systems can compare location with the expected runway, and surface-detection systems can generate conflict or wrong-surface alerts at equipped airports. None of these technologies is infallible, but an independent alert can interrupt a sequence when both pilots share the same misconception. Construction risk assessments should similarly test whether an unfamiliar crew can interpret the temporary layout correctly in darkness, poor visibility, and high workload.
A Systems Interpretation of the Accident
James Reason’s “Swiss cheese” model provides a useful framework for understanding Flight 5191 because it describes accidents as the alignment of weaknesses across multiple defensive layers rather than the product of one isolated failure (Reason, 1997). In this event, potential defenses included taxiway design, construction information, runway markings and lights, cockpit attention, sterile-cockpit discipline, checklist behavior, heading verification, controller observation, and surface-warning technology. The immediate unsafe act was the decision to begin takeoff from Runway 26, and the crew remained responsible for verifying its position. A systems approach does not erase that responsibility. Instead, it asks why the mistake survived long enough to become fatal and how the system can make the same error less likely for another competent crew. Replacing the pilots without changing procedures or defenses would not guarantee prevention. Effective safety improvement strengthens several layers simultaneously so that an error in one layer encounters a barrier in another before the aircraft reaches an irreversible state.
Safety Lessons and Prevention
The strongest preventive lessons from Flight 5191 concern explicit verification, protected attention, temporary-airport risk, and independent alerts. Airlines should require a verbal runway-and-heading cross-check immediately before takeoff, with both pilots independently confirming runway identity from signs, heading, charts, and available electronic displays. Any mismatch should create a mandatory stop condition. Scenario-based training should include wrong-runway traps, construction layouts, darkness, time pressure, and the social discomfort of admitting uncertainty. Airports should evaluate temporary markings and taxi changes from the cockpit perspective, not merely for formal compliance, while controllers should be supported by staffing and procedures that preserve attention during active movements. Technology should be deployed according to risk and designed to provide information independent of the crew’s existing assumption. The broader principle is simple: aviation should make the cautious choice easy. A delayed departure creates inconvenience, but beginning a takeoff with unresolved position uncertainty can turn a routine navigation error into a catastrophic event.
Conclusion
Comair Flight 5191 demonstrates why aviation accidents cannot be understood adequately through the phrase “pilot error.” The pilots did enter the wrong runway and failed to verify their position before takeoff, but the safety value lies in understanding how that error survived several opportunities for detection. Expectations shaped perception, nonessential conversation competed for attention, construction altered the surface environment, the dark runway did not trigger a stop, controller observation did not provide a final barrier, and no automated surface-warning system interrupted the sequence. The NTSB’s probable-cause finding appropriately focuses on the crew’s failure to use available cues and cross-check runway identity while also identifying contributing conditions. Prevention therefore requires both professional accountability and system design. Checklists, runway-heading calls, sterile-cockpit discipline, construction risk assessment, surface surveillance, and cockpit awareness technology can reinforce one another. The enduring lesson is that uncertainty should produce verification before commitment, because a high-reliability system must prevent one mistaken assumption from becoming the final decision.
References
National Transportation Safety Board. (2007). Attempted takeoff from wrong runway: Comair Flight 5191 (AAR-07/05).
Reason, J. (1997). Managing the risks of organizational accidents. Ashgate.
Academic Master Education Team is a group of academic editors and subject specialists responsible for producing structured, research-backed essays across multiple disciplines. Each article is developed following Academic Master’s Editorial Policy and supported by credible academic references. The team ensures clarity, citation accuracy, and adherence to ethical academic writing standards
Content reviewed under Academic Master Editorial Policy.
