I've been recording for a while now at 24bit 44.1khz. Can anyone actually hear a difference with higher sampling rates? It seems to me that  at higher sampling rates, your average listener in the modern world (high end topping out at 15-18khz) wouldn't be able to even hear it. Nyquist seems to suggest that 44.1khz gets you up in the near 20khz range ( although you're not getting more than a couple of sample points @ 20khz). What say you? I know I may be opening up a can of worms and this has been talked to death, but I'm seriously considering it for some projects and could use some input.

From the little I understand I think that the higher sample rates give you more headroom while tracking and using VSTs and whatnot. When you bounce the end result is supposed to be better I guess. I'm sure someone more knowledgeable can enlighten us however.

Higher sampling does not affect headroom. Bit depth determines headroom. Higher sampling rates may reduce sampling artifacts. Some harmonically rich VSTi patches may produce audible artifacts at 44.1. This depends on the synth and patch.
For VSTi's, Sonar has the ability to apply oversampling (for some VSTi's at least) to the arithmetic calculations producing the synth wave.
No need to increase project sample rate in this case.
There is much controversy regarding higher sample rates for mastered (final) material. Most double blind listening tests fail to show discernable improvements above 44.1.
Be aware that increasing sample rate increases the load on the PC (CPU, disk, memory).
It is up to you to decide if it is worth it. I use 48/24 or 44.1/24.
 
 
 

A zillion threads on this exist...it's a never ending subject of discussion...just Google and spend your weekend reading...regurgitating the same stuff is a waste of time as nothing has changed.

Most of us are at 44.1 or 48/24 and that's what cakewalk recommends. I've tried others...higher reduces latency with more cpu load.

Read the other threads...

Vastman
Most of us are at 44.1 or 48/24 and that's what cakewalk recommends. I've tried others...higher reduces latency with more cpu load.

48 will lower your latency relative to 44.1 for a given buffer size, however most audio interfaces I've seen double their buffer size for 88.2 or 96 negating any gains.

Here is my rule of thumb...for HD stuff e.g. classical, acoustic and some...other 96khz is great...there is air and sparkle...hence HD stuff...for everything else...44.1khz is fine...oh and sound effects, sound art and sonic art type stuff because it might be used on Youtube or something 96Khz is what you would want too.  Oh Oh and I forgot 96khz is great for time based effects, I can't remember off the top off my head the reason, but I know it has something to with reverb tails.  And time based effects are chorus, flanger reverb and delay. 
 
Maybe try an experiment, do a track or a recording at 96khz, bounce it down and see what it sounds like on a normal system, then find a 44.1khz track that you've recorded and see if you can hear the difference.  This experiment should be electronic, and acoustic IMO.
 
Peace Ben

It is thoroughly explained in Ethan Winer book "The Audio Expert". Together with other audio myths.   http://ethanwiner.com/book.htm 

I use 48 most of the time.

I use 44.1/24 all of the time

48/24 24/7.