Hearing loss is routinely estimated from the audiogram, even though this measure gives only a rough approximation of hearing. Indeed, cochlear regions functioning poorly, if at all, called dead regions, are not detected by a simple audiogram. To detect cochlear dead regions, additional measurements of psychophysical tuning curves or thresholds in background noise (TEN test) are required. A first aim of this study was to assess the presence of dead regions after impulse noise trauma using psychophysical tuning curves. The procedure we used was based on a compromise between the need to collect reliable estimates of psychophysical tuning curves and the limited time available to obtain these estimates in a hospital setting. Psychophysical tuning curves were measured using simultaneous masking with a 2-alternative forced choice paradigm, where the target was randomly placed in one of the two masker presentations. It is well known that some components of noise-induced hearing loss are reversible. A second aim of this study was to examine the potential recovery of dead regions after acoustic trauma. A third issue addressed in this article was the relationship between noise-induced dead regions and tinnitus. We found that 70% of the subjects had dead regions after noise trauma, while 88% reported tinnitus. Moreover, we found that the extent of dead regions probably diminished in about 50% of subjects, which highlights the ability of the human auditory system to recover from noise-induced hearing loss.