You can also bring the audio into something like Audacity and check the time numbers. Then, divide each second into fifteen parts, (for 15FPS) write down when what needs to be doing what when, and you've got a frame-by-frame guide that should be right on the dot. Not only that, but you can easily see how many frames each movement will need, thus helping with the easing in/out.
Is that clear enough?

You can just import the audio into a monkeyjam project, that makes it really easy to see how many frames you'll need for each syllable/word or whatever.

BRAWL 2013 ENTRY Quack In Time
"Why in the world did you do a weird language if you know English?" - tenny1028

I think you should:
1 make the audio
2 animate as best as you can
3 adjust the audio so that fits into the animation

When in doubt, animate more. it's better to have too much than too little. Also have a pose that your minifig returns to after each arm or head movement.  If necessary, you can also have the same action happen twice, although preferably not back-to-back. If your still having trouble, see if there are any unnecessary words in your audio. That's all I have

Unfortunately math is the only way to line up animation to audio precisely. Fortunately, it's actually really easy once you understand why you're doing it. Prichard got a little turned around though (or I just don't understand the technique he's using): I think it's easiest to multiply the position in the audio track by 15. When you're doing 15 frames per second, each frame stays on the screen for 1/15th of a second. So the beginning of a one second sound clip is at frame zero (0*15=0), and the end is at frame 15 (1*15=15).

Say you had a line, "Why would you do that?" that took 3.2 seconds, and you wanted to emphasize the word "that?" (suppose "that?" starts 2.75 seconds in and ends at 3.2 seconds). In your animation the word "that?" would start at (2.75 * 15 = 41.25) the 41st frame (because you can't have .25 of a frame, I've just discarded the remainder) and it would end at (3.2 * 15 = 48) the 48th frame. So you know you need to fit whatever motion you want to use to emphasize "that?" in the 7 frames between the 41st and 48th frame in your animation. It sounds complicated, but it really isn't. I used to use a tool called WavTRACKER which lets you select parts of your audio and automatically gives the frame equivalent for the selection. It's a pretty simple tool, but can speed things up a fair amount- I'm sure there are other programs that serve the same purpose if you don't want to convert to WAV files or use a non-Windows machine.