Yes, what 2 pass does is it goes through the video ones (to examine it) then a 2nd time to encode (and find more optimizations)
Well, I really over simplified it, but that's the gust. :)

2-pass should be better, but in practice you may not be able to tell much difference if you are already encoding at a fairly high bitrate (like 7,000 and above). The reason for this is that the optimizations done by the first pass pinpoint fast-moving segments of video that need higher encoding rates to avoid compression artifacts. However, if the difference between the average bitrate and the maximum bitrate is relatively small, you won't see much difference, no matter how much optimization is performed. The extreme example would be encoding at an average of 8,000 with a maximum also set to 8,000.

The implication of the above paragraph is that 2-pass is very important if you are forced to encode at low bitrates (anything below 6,000, is my definition of low bitrate).

Also, if you have video that has the same amount of motion in every frame (like a television interview show), then no matter how many times the algorithm searches, it won't find video that can be encoded at lower bitrates to balance fast moving video. Most project usually have some fast moving video, but many training films and interviews do not, so 2-pass is probably a waste of times with this sort of program material.