Currently I've been using a certain chart to outlay the basic contents of a dungeon (then I go and modify the overall result slightly, to fit how I see the creatures having gotten there and living there). The chart also determines treasure.
1-2 = Monster, 3 in 6 chance of treasure
3 = Trap, 2 in 6 chance of treasure
4 = Trick, 2 in 6 chance of treasure
5-6 = Empty, 1 in 6 chance of treasure
But I want to get the most out of each session in the time given. Granted, sometimes the players see a room with monsters in it, but no treasure - but rush them anyway! But really, I'm not quite interested in these sorts of rooms and the RL time it takes to run them.
So what if we assumed we skipped all the rooms that had no treasure in them?
Looking at the chart, you could take the odds of treasure and make a new chart - monsters having a 3 in 6 chance of treasure and there being a 1-2 chance of monsters, lets say 'monster with treasure' has a 2*3=6 point chance of occuring. Do that to the rest of the chart, then make it work on an appropriate die and you get...
1-6 = Monster with Treasure
7-8 = Trap with Treasure
9-10 = Trick with Treasure
11-12 = Empty with (hidden) Treasure
So roll 1D12 on that chart!
PS: Of course empty rooms aren't just empty - if something comes to mind for what could be in it (ie, maybe it stores food, or is a prayer room) then that's what's in it.
Edit: On reflection I would change the trap and trick to continue to be 2 in 6 chance of having treasure. Because while you might be able to see a room has a monster but no treasure, you can't exactly tell if a room with treasure is trapped or not, so you couldn't skip those. With the empty rooms, they continue to have hidden treasure, which isn't obvious because often trapped/tricked rooms without treasure look like empty rooms as well.