Recently, early childhood educators and policy makers have placed an increased emphasis on assessing emotional development in young children who are considered educationally at-risk. (Department of Health and Human Services Administration on Children, Youth, and Families/Head Start Bureau, 2000). The desire to assess emotional development is not surprising given that children's skill in understanding others' emotions appears closely tied to their ability to establish and maintain positive social relationships with others in the school years and throughout life. Specifically, researchers have found that young children who have better developed emotion recognition and emotional perspective-taking abilities demonstrate higher levels of prosocial behavior and are more popular with their peers (Denham, 1986; Denham, McKinley, Couchoud, & Holt, 1990; Leppanen & Hietanen, 2001). Other research has linked emotion recognition and understanding with high levels of pretend play and behavioral and emotional competence with peers (Carlo, Knight, Eisenberg, & Rotenberg, 1991; Lindsey & Colwell, 2003). In contrast, children who are poor at emotion recognition and emotional perspective-taking are at increased risk of being disliked by their peers (Denham et al.; Leppanen & Hietanen, 2001), are rated by teachers as being more socially withdrawn (Strand, Cerna, & Downs, in press), and are at risk for the numerous negative educational and psychosocial outcomes long known to be associated with peer rejection (Asher & Wheeler, 1985; Asher & Coie, 1990; Ladd, 1990). Although results such as these have suggested that emotion understanding in general, and emotion recognition in particular, are important developmental variables, there is limited information available regarding the effectiveness of attempts to manipulate emotion recognition skill in individuals who have disabilities. Research with adults has generally indicated that interventions can successfully improve the emotion recognition ability of those who have learning disabilities, (McKenzie, Matheson, McKaskie, Hamilton, & Murray, 2000), intellectual disabilities (McAlpine, Singh, Ellis, & Kendall, 1992; Rydin-Owen, Drake, & Bratt, 1999), high-functioning autism (Bolte, Hubl, Feineis-Matthews, Dierks, & Poutska, 2006), or acquired brain injury (Guercio, Podolska-Schroeder, & Rehfeldt, 2004). However, studies focused on children have been less consistent. For example, Dyck and Denver (2003) reported no improvements in emotion recognition skill for deaf children exposed to an 11-lesson psychoeducational program. In contrast, Stewart and Singh (1995) found that teaching via directed rehearsal led to significant gains in basic emotion recognition skill in a study of boys with mental retardation. Emotion recognition has long been thought to be impaired in children with autism spectrum disorders (ASD; Hobson, Ouston, & Lee, 1989), so it is not surprising that some researchers have examined the effectiveness of emotion recognition training with this population. As has been the case in children with other disabilities, the results have been equivocal. Using a computer training program, LaCava, Golan, Baron-Cohen, and Myles (2007) found significant gains in emotion recognition skill in children with ASD. However, Silver and Oakes (2001) reported that computer training did not lead to significant emotion recognition gains in adolescents with ASD compared to controls. Other research has shown that children with autism exposed to behavioral interventions may develop advanced emotion understanding abilities (e.g., the understanding of desire-based emotions in others) but may still have difficulties with basic emotion recognition (Downs & Smith, 2004). It is interesting to note that none of the studies cited above attempted to teach emotion recognition skills to preschoolers, even though the developmental literature suggests that the preschool years are typically a time of significant growth in this ability (Denham, 1998). …