An important problem in cognitive and systems neuroscience concerns the extent to which perceptual organization can be explained by “local”, peripheral physiological mechanisms, or rather by more “global”, central, and higher-level processes. Though central in vision research, this issue has received little attention in the field of audition. One claim is that auditory-perceptual grouping mechanisms, possibly related to visual figure-from-ground segregation or “pop-out”, are low level, resulting from local processing in the frequency domain. However, no experiments have been performed specifically to test this question. We examined the effects of perceptual grouping on detection for reversal of two repeated target tones, one constant in frequency (1030 Hz), the other free to vary between trials (1045-8580 Hz). Detection was examined in the presence of a 1000-Hz background tone that repeated between target presentations. By varying the frequency of the high-target tone, this task was designed to modulate grouping between the background and low-target tones, thereby affecting reversal detection. We predicted that at large target frequency differences (Δf), the high-target tone would segregate from the background and low-target tones, and so render the background and low-target tones less distinct. We found that reversal detection declined from optimal levels with increasing Δf, and that performance was improved by spatially separating the location of the target and background sounds by at least 32°. These results demonstrate that global frequency integration over at least three octaves occurs through grouping, and that grouping is affected by source location. This implies that auditory-perceptual grouping involves global neural processing, i.e., the participation of neurons with very broad frequency input that are also sensitive to spatial location.