Rather than devolve into some kind of QA fiesta, I'll just relink my original entry thread here:viewtopic.php?f=23&t=988
It pretty much details the process almost day-by-day from start to finish (and post-launch).
Keep in mind though that everyone has their own mapping methodology, their own target goals along the way, and their own opinions about what should be prioritized. I spent most of the competition doing geometry work and hacked together gameplay at the last minute, but it seemed to have worked out at least mostly OK for the initial run (aside from the impossible-to-locate climbable chains, haha).
I was also greatly advantaged by having a complete map plan right from the start and self-imposed constraints (no indoor geometry, lots of climbing, heavy combat) that I used to keep myself focused. If you can do that, you'll be a lot better off, but you'll need a clear idea of where your map is going right from the start, or at least a very clear idea of what the setting will be composed of.
I would recommend trying to be done with basic geo by the half way mark and spend the rest of the period optimizing and doing gameplay/story/detail work. You WILL miss the mark for geo because UE1 hates you, so if you set the mark farther ahead, you'll be better off when you wind up having 2 weeks left in the contest for everything else instead of just 2 days.