I constantly hear that air accidents are caused by a range of shortcomings. However, most accident reports declare a main "probable cause". You can see below a bar-chart with the causes of fatal accidents by decade (data source):

enter image description here

Does anyone know which criteria are used to determine the main accident cause?

  • 1
    $\begingroup$ The page you linked actually give the criteria for each of those buckets. Is anything specific unclear? $\endgroup$
    – Jamiec
    Jan 29, 2020 at 10:09
  • $\begingroup$ @Jamiec Indeed, the page breaks down the categories into specific causes, but it does not provide information about how is this cause chosen. For instance, let's say pretend that an air accident involved a design flaw and the crew executed an improper procedure. Eventually, the accident cause was reported as "Mechanical". My question here is: Why "Mechanical" and not "Pilot Error"? $\endgroup$
    – ppinto
    Jan 29, 2020 at 10:17
  • 4
    $\begingroup$ The falacy here is that there is one cause of an accident. They are always a "chain-of-events" which can be broken at any time, and no accident occurs. I think what were talking about here is the major contributing factor. I will try to find a decent answer, unless someone else does here. $\endgroup$
    – Jamiec
    Jan 29, 2020 at 10:20
  • 1
    $\begingroup$ Please note that the categorization you talk about seems to be specific to that specific website - not something that is widely used in aviation in general. So to get the best answer you should probably contact the owner of that website. $\endgroup$ Jan 29, 2020 at 10:41
  • $\begingroup$ @J.Hougaard Do you know the standard terminology for air accident causes? Maybe is it different among each Civil Aviation Authority? $\endgroup$
    – ppinto
    Jan 29, 2020 at 11:21

1 Answer 1


I participated in a tiny supporting role in a few crash investigations involving CRJs, so I'll give it a shot, and I would say that when a main cause is listed, especially for something mechanical, it's what you might call a "primary trigger", without which, follow-on supporting factors and events that led to the crash would never have occurred. Or an event in a chain that can be considered the "point of no return".

Pilot with poor training takes off in an overloaded airplane, and engine quits after V1/VR and it crashes. The engine failure will be the main cause, with poor training and overloading added as contributing factors. Without the engine failure, the poorly trained pilot in his overloaded airplane would've carried on his merry way, so the engine failure can be considered the main cause.

For pure "pilot error" events, it can be more nuanced, with a roughly equal-in-impact chain of events and if you took out any element in the chain, the other elements may have still happened, but just not leading to a fatality. These reports will often list a series of accident causes without giving critical weight to any one, except to say that the cause was ultimately the crew and not the plane. Running out of gas is a typical one.

On the other hand, sometimes a crew action is as much a triggering event as a major mechanical failure. Colgan Flight 3407 is a classic example of this. The Capt had the Mother of All Brain Farts and pulled when he should have pushed when the shaker went off and this is listed at the main cause. Now, the shaker went off in the first place because they weren't paying attention to power when the A/P leveled off, but it was the response to the shaker that sealed their fate, and becomes the main cause.

There was a CRJ crash in China caused by taking off with frost on the wings on a cold clear morning. Frost on the wings was the main cause. Lack of understanding of the effects of frost, and maybe a bit of normally benign over-rotation-on-takeoff contributed, but without the frost, nothing would have happened.

  • $\begingroup$ This is a very interesting point based on your personal experience which I appreciate a lot. I will not mark it as resolved yet hoping that someone else contributes with a different perspective. $\endgroup$
    – ppinto
    Jan 30, 2020 at 16:27
  • $\begingroup$ No worries. You'll might get lucky and find someone who has direct experience writing or reviewing Safety Board reports (I provided indirect technical input to my old employer's Product Safety team). If I was still working I could go chat with a Product Safety guy at my old employer and ask directly, but I retired several years ago. $\endgroup$
    – John K
    Jan 30, 2020 at 19:33

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .