When I first started playing with MP3, I thought it was lossless compression, sort of like WinZIP for audio. Boy was I wrong, MP3 is a lossy format, meaning it has to throw away information based on a psychoacoustic model to which parts of the sound you can hear, and the ones you can't. This is why such small filesizes can be achieved with lossy compression, however, artifacts can be inherent and/or annoying to you if the psychoacoustic model is incorrect, or the bitrate doesn't permit the artifacts to be hidden. Say you rip a song from a CD, keep the original, and encode it into an MP3 at 128Kbps. Then you decode the MP3 into a WAV file (notice it's not referred to as decompressing the MP3). Those two WAV files will be damn near the exact same size, but they will sound different. This is because the MP3 encoder had to throw away data that cannot be recovered from the MP3 during decoding. So, consequently, this means that if you re-encode an MP3, there will be further quality loss than if it was encoded from the original source.
But anyways, the simple answer is if you encoded two seperate MP3s at 128Kbps and 192Kbps, then decoded them, they will be the exact same size WAV files.