1) About 25% of the questions were exact repeats of questions on the 2003 exam.
2) There were at least 3 questions where reasonable people could disagree on the correct answer.
3) This particular version of the exam was heavy with questions on Appeals and 103/102 rejections. Also, there were at least five PCT questions, three of which featured the same fact pattern, with the citizenship of the inventors changed.
4) I only spent about 5 days studying (about 4 hours per day), figuring that I could make up for my lack of study with my "real world" experience as a patent prosecution associate at my firm. I took some practice tests at home, and bombed on all of them (generally only getting about half the questions right). I went in today thinking I would have to retake the exam later this year.
Well it turns I was right about the experience, but for the wrong reason. Where my experience came in handy was that I knew exactly where in the MPEP to look for the answer, and I looked up about 75% of the questions.
5) I understand patent bar passing rates are somewhere in the 50-60% range. That's pretty low, relative to state bar passing rates in most jurisdictions. I suspect the reason for that is that (a) you don't get to take a semester long course on the MPEP, and (b) the MPEP is written in such a non-intuitive way.
6) Is this exam really the ideal way to determine competence/expertise to prosecute patents? The technical degree requirement for registration essentially eliminates anyone who doesn't have a minimum level of technical competence. So, as far as patent prosecution competence, wouldn't it make more sense to have a 6-hour exam that featured claim drafting and/or amendment, and responding to Office Actions?