Unquestionably. I once did a personal test using Foobar 2000 to calibrate myself about the differences from hi-res vs 44.1/16 and vs perceptual encoding.
The test plug in had an ABX blind tester, but I first wanted to try it unblinded to train myself.
The test started with a classical 88.2/24 file (I listen to nearly all classical/ambient and go to live concerts with reasonable frequency), resample to 44.1/16 then perceptually compress, I think either 320kbps MP3 or 256 kbps VBR AAC. Then convert everything back to 88.2/24 and feed back to the DAC. The reason for the second upconverting is that DAC's can be different at different sampling rates because of their internal design, and that should be eliminated.
The result was that through my DAC/AMP and headphones (HD580/D5000) even sighted I couldn't convince myself I heard any substantial difference between any of them if I was just listening to music normally. I didn't bother with the ABX test or distinguishing 88.2/24 from 44.1/16
Perhaps if I had concentrated on a tiny segment of music, probably one with lots of non-overtone high frequencies (e.g. percussion/cymbals) which is difficult to compress, then maybe there could be a difference. The sound quality differences between different recordings (say in MOG listen to 10 different recordings of the same classical piece) is enormous.
I decided to not worry about perceptual coding at 256kbps and up, any change was well, well below my threshold for caring. Not coincidentally this bit rate (or below it even) is where scientific perceptual research shows that almost all people stop being able to distinguish uncompressed from compressed in blinded tests.
128kbps constant-bit-rate MP3 is noticably worse and 'flatter'---not in the frequency response sense but the depth sense.