The goal of our research was to determine whether there are gender differences in nonliteral language use when people communicate about emotions. In Experiments 1 and 2, participants watched characters in film clips experience emotional events. In Experiment 3, participants read narratives that were analogous to the film clips. Later, they wrote a description of the character’s emotion or a description of how they would have felt in the same situation. There was no correlation between the perceived intensity of the emotion and the amount of nonliteral language in the participants’ descriptions, which is problematic for Ortony’s (1975) vividness hypothesis. Men used more nonliteral language in descriptions of negative emotions than positive emotions, whereas no difference was found for women. Finally, men tended to use more nonliteral language in descriptions of others’ emotions, whereas women tended to use more in descriptions of their own emotions. These findings suggest that men and women do use nonliteral language differently, at least in the context of emotional communication.