This dissertation investigates how segmental contrast and prosodic structure are acoustically implemented in production and processed from the acoustic signal in perception. In Seoul Korean, it has been widely accepted that the primary acoustic cue for signaling prosodic structure is F0 contour (Jun, 1993). Prosodic structure has also been reported to affect the voicing of lenis obstruents in Seoul Korean (Jun, 1993) but due to the finding that this segmental process is optional (Jun, 1994; Han, 2000), it has been assumed that segmental realization is a less reliable cue for prosodic structure than the F0 contour (Jun, 1994). This dissertation presents findings that suggest that there may be a greater role for segmental realization in signaling prosodic structure than previously thought. First, I report empirical evidence that suggests the realization of lenis obstruents is in fact more systematically affected by the prosodic context than previously reported. Second, contrary to previous assumptions that F0 contours more reliably signal prosodic structure than segmental realization, I report evidence that segmental realization may in fact be more reliable in signaling prosodic structure in certain contexts in Seoul Korean. Finally, these findings that the realization of lenis obstruent is a reliable cue for prosodic structure are supported by the results of a perception experiment, which shows that listeners are more affected by the segmental realization of lenis obstruents than by the F0 contour in making decisions about prosodic structure. Based on these findings, I argue that acoustic cues simultaneously encode both segmental contrast (lenis vs. aspirated stops) and prosodic structure (phrase-initial vs. phrase-medial) in Seoul Korean, and listeners hearing these acoustic cues decode both types of information simultaneously.