Two experiments were conducted on musical instrument identification. The first experiment investigated performance of normal-hearing (NH) listeners in a closed-set musical instrument identification involving vocoder simulations of a cochlear implant (CI) and electric acoustic stimulation (EAS; where a hearing aid supplements the CI to access low-frequency residual hearing). The second experiment investigated performance on the same task by bimodal CI users who normally wore a hearing aid on their contralateral ear. While the NH listeners exhibited an EAS benefit over the CI condition for particular instruments, bimodal CI users did not show a consistent benefit of EAS over CI alone and exhibited the highest performance when listening through their hearing aid alone. This finding suggests that there may be factors (such as duration of device use and interaural frequency mismatch) that may currently limit the EAS benefit for music listening in the bimodal CI population. The identification rates varied across the different musical instruments, with higher scores for instruments with more rhythmic excerpts. This finding supports previous reports of the relatively greater salience of temporal envelope and rhythmic cues over other timbral and pitch-based cues in CIs. The comparison of the results of the two experiments emphasizes that predictions based on simulations may not always translate to clinical outcomes in the CI population.