Personally, I have a couple of different stages of sharpness that I use to evaluate.
First stage, hold the knife end on and look down the edge. If you are at a reasonable starting point, then the edge should look like two lines meeting. This can also help with the sharpening process, because you can often see if you are hitting the bevel instead of the edge when you sharpen, you can also see if you are raising a burr as well.
Second stage, run my thumb across the blade edge (not down the length) to see if the edge catches in the ridges of my thumb print.
Third stage, shave hair - going both directions and with either hand using the knife.
Fourth stage, a couple of different tests. Can I:
Slice down the edge of a piece of regular printer paper and create a thin long curl? (You can substitute the paper shivving test described above as well here. Periodically I do with a thin blade.)
With a quick slice while holding the paper in the air slice a section off of the paper, without tearing it?
Take a long slow slice through the paper traveling the full length of the edge to see if it slices cleanly along the entire length?
Fold the piece of paper in half, have it stand on a cutting board, in an angle, looking like an L from the top and use the knife to slice down the length of the fold without crushing the paper.
I've also used the thumbnail test described above. Incidently, not every edge geometry would pass all of the tests described. For whatever reason, I have the worst time with my Benchmade mini - stryker. I'm too lazy to reprofile it (yet) but it is by far the worst cutting knife I own.
