There is a strategy called event counting. By relating the bottom edge of the top view with the top edge of the front view, and the top edge of the end view with the right edge of the top view and counting their "events", you can eliminate most of the wrong answers. Here is an example:
So looking at the top edge of the end view, I count all of the "visible events" represented by a red circle and the "invisible events" represented by a red square. I know that this edge corresponds to the right edge of the top view:
Now, by comparing that information with the answer choices, say B, I can eliminate some of them. The blue markings show an incosistency with the relationship that we made in the previous step.
You can do the same thing with the top edge of the front view and the bottom edge of the top view.
By comparing this relationship with the answer choices , you can eliminate the rest of the wrong answers, like A shown below.
In just about every case, this process will eliminate all of the wrong answers. It is every helpful when you can't visualize the actual object. All you have to do is look down the edges and compare them. It almost never fails, but if you are going to use this strategy, you should practice lots and be sure that you know how to do it correctly or it will eat up a lot of time.
I am to the point where I can visualize the three dimensional objects for most of the problems just by seeing the two views, but this strategy never fails when I have a problem. Good luck!