Param -
Conventional Wisdom puts the number at less that 25% but as Rebecca has stated that is not publicly available.
However, good test strategy does not depend on knowing the number. Since we know that some questions do not count - say 1/5 or 1/6 are "unscored items" - we know that spending more than 4 minutes on any question really is a bad idea since even if you get it right there is a reasonable chance that it does not count! The presence of these unscored items just helps emphasize a few things:
1) do not get caught up on any one question. Even if you get it right it may not count and spending too long on any one question can hurt your score anyway.
2) NEVER try to figure out how you are doing during the exam. If you base your idea of how you are doing on some "easy" question that you got - or some "hard" question for that matter - it may have been an unscored item that is not even adapted to you. Also, it is really impossible to know just from reading a question what percentile it would be. Simple for you could be tough for someone else.
Speaking of "Easy questions" here is an article I just wrote about E A S I questions
https://www.beatthegmat.com/mba/2012/12/ ... n-the-gmat
And here is a fun one on GMAT strategy
https://www.beatthegmat.com/mba/2011/01/ ... n-the-gmat
