I got a 48 on my most recent test, and I was playing catch-up by 10th question - that was a huge mistake. GMAT will ask you concept questions that you simply cannot cover in the prep, or ask you in a way that you won't be able to figure out unless you're just that darn good at quant or backsolving. It's hard to give up, but it's VERY important that you must cut your losses.

There was one question toward the end that I know I could solve within 2-3 minutes, but I only had time for 90 seconds per question left and I knew that question cannot be solved under 90, so I had to move on. It's rather interesting that I would know when to move on for a question that I know answer to, but I stumbled for 4-5 minutes on a concept question that I really couldn't fully grasp. In retrospect, I should have skipped that concept question and worked on this one instead, and maybe my quant would have gone up another point.

better to cut your losses in the middle of the test and have a strong finish.
middle questions have a high chance of being experimental then then last few of a section.

Also, I rather to take the "blind" guess (1/5) on a very hard PS question, and use 60seconds to narrow down a DS question to 2 or 3 choices then guess. You can easiler narrow down a DS question then a PS question, even if you dont understand the question/statements.

I got a Q49 and I had to guess on at least 3 questions. Usually you can look at a question and know if you are going to struggle with it or not. On those, its probably best to cut your losses and just guess. The only thing I will say is make sure you get all of the easier questions right.

I recommend the following strategy as regards guessing and timing strategy:

1) Focus on getting questions right that you can get right. In other words, have a strategy that helps you to avoid errors such as answering the wrong question, making assumptions, and calculation errors. Many people try to "make up time" on the "easier questions" and wind up missing some of those. This you cannot afford. As Osirus says, make sure you get the easier ones right.

It seems that the more a student has studied and the better they are at the quantitative portion, the greater the proportion of their mistakes are these "avoidable" errors. Bottom line, if you can get a question right - you must commit the time and the focus to do so.

2) As the Duke and UWHusky discuss, you have to know when to let go of a question. 4 minutes spent early on in the test to work on a problem (whether or not you eventually get it right) will result in tears shed as you have to quickly guess at, rush on, or worst of all not even get to answer some question at the end of the test, a question that you could absolutely get right in 2 minutes.

I recommend that if you do not have a firm strategy for solving a problem in about 1:15 seconds, it is very likely that you will not be able to answer the question in less that 3 minutes. So better to guess and move on.

Remember this about guessing, with so many experimental questions on the exam - perhaps 1 out of 5 questions (or more) you have around a 20% chance that the question you are guessing on will not count. You also have a 20% chance of guessing the correct answer. Of course these two categories do overlap, but you will still have something like a 35% chance that a question you guess on will not count against you. And remember that if after 3 minutes you have managed to narrow a quant question down to two choices that you view equally your chances of getting that question right - and having it count in your favor - are not as good as 50% (there is that chance that you guess correctly and it does not count) 3 minutes wasted!

You should try to have only two categories of questions on the test: those questions that you devote yourself to getting right and those questions that you recognize will be potential time sinks that you guess on and move away from.

Hope that helps!

That helps a lot! Thanks everybody!! Appreciate it a lot.