Vocal emotion production varies across talkers. Effects of talker variability (i.e., same talker vs different talkers across trials) on vocal emotion recognition were tested in babble noise and with cochlear implant (CI) simulations in normal-hearing listeners. Results showed significant talker variability effects in noise but not with CI simulations, suggesting that the degree of talker variability effects on vocal emotion recognition may vary with listeners’ weighting of acoustic cues for emotions (e.g., pitch and duration) and the availability of these cues in different listening conditions. Vocal emotion recognition significantly improved with higher signal-to-noise ratio and more vocoder channels.