1) Delegate
Graduate students were invented for a reason. Delegate everything. Data analysis. Literature reviews. Writing. Teaching prep. If you don't have any graduate students, find some undergrads. A task you might find mundane can actually be "research experience".
One of my colleagues laments that their students often mess things up, causing even more work than just doing it themselves. For some things this may be true, but thus far I've been pleasantly surprised with what happens when I challenge my students. Even if they don't do a perfect job they are still doing useful work for both of us.
For CS-type folks -- get someone else to manage your machines. I know it's tempting to spend hours and hours tweaking your linux box to play fur elise backwards every other Tuesday at 2:22pm and building That Perfect Windows Manager config file and whatever other fun hacky things you do, but try to resist the urge. Shell scripts won't get you tenure (damnit).
2) Ask
I read Boice's advice for new faculty, and he implied "quick starters" ask everyone for help on everything - research, teaching, grant writing, etc. This is helpful to remember. Spending three hours trying to find an arcane policy statement on NSF's website isn't worth the time when a quick email to your university's research support office will suffice.
3) Network
Don't forget to water your social networks, both in your department, at your university, in your town, and in your professional community. This is something you shouldn't delegate or put off.
Of course you don't want to spend all of your time networking; I think Boice suggests about 2-5 hours a week. YMMV.
4) Dry Clean
Best. Thing. Ever. Absolutely worth every penny. To just wake up in the morning, go to your closet, have cleaned and pressed professional clothes ready to go is just about the best thing since fur elise.
------
(*) To be honest I just had one piece of advice ("Dry cleaning FTW!"), but I thought I'd throw the other ones in too while I was here.
