Posed emotional facial expression was examined in brain-damaged adults with right (RBD) or left (LBD) hemisphere cerebrovascular lesions and in matched normal controls (NC). Subjects were videotaped while posing eight facial expressions (positive and negative) under two elicitation conditions (verbal command and visual imitation). Expressions were rated by four naïve judges for intensity, category accuracy, and valence accuracy. RBDs were significantly more impaired than LBDs or NCs on category and valence accuracy, while LBDs posed expressions with significantly more intensity than RBDs or NCs. These findings held for positive emotions only. The results for category accuracy replicate an earlier report (Borod et al., 1986) which utilized a different rating procedure. Finally, expressions posed to visual imitation were rated as more intense and more accurate than those posed to verbal command.