Young adults with no history of hearing concerns were tested to investigate their /da/-evoked cortical auditory evoked potentials (P1-N1-P2) recorded from 32 scalp electrodes in the presence and absence of noise at three different loudness levels (soft, comfortable, and loud), at a fixed signal-to-noise ratio (+3 dB). P1 peak latency significantly increased at soft and loud levels, and N1 and P2 latencies increased at all three levels in the presence of noise, compared with the quiet condition. P1 amplitude was significantly larger in quiet than in noise conditions at the loudest level. N1 amplitude was larger in quiet than in noise for the soft level only. P2 amplitude was reduced in the presence of noise to a similar degree at all loudness levels. The differential effects of noise on P1, N1, and P2 suggest differences in auditory processes underlying these peaks. The combination of level and signal-to-noise ratio should be considered when using cortical auditory evoked potentials as an electrophysiological indicator of degraded speech processing.