Perception of dichotic chords (free recall and directed recall), nonsense syllables (CVs), and three-pair digits was assessed on 24 musicians and 24 nonmusicians. On the dichotic-CV and dichotic-digit free-recall tasks, there was a significant right-ear advantage, but there were no group differences. With the dichotic-chords, free-recall condition, a significant left-ear advantage was observed but no group difference. For the dichotic-chords, directed-recall conditions, the musicians performed significantly better by 10 percent than the nonmusicians. Unexpectedly, for the dichotic chords, the 62-72 percent correct performances were better on the free-recall condition than the 42-55 percent performances on the directed-recall conditions. These differences between the two response modes were attributed to the difficulty of the dichotic-chord listening tasks and the probabilities associated with the closed-set response paradigms. The findings suggest that the dichotic-chord paradigm used in this study should not be included in clinical protocols used to assess auditory perceptual abilities.