It is widely assumed that (1) children with challenging behavior may lack the social–emotional skills to be successful in classroom settings and (2) instruction in social–emotional skills can help prevent the occurrence of young children's challenging behavior. However, no data are available to indicate whether specific social–emotional skills might be most effective in reducing existing challenging behavior. In this study, we used an archival database to examine possible relationships between types of social–emotional instruction and changes over a 4-month period in levels of social skills and challenging behavior. The database, from a randomized trial of an intervention for preschoolers' challenging behavior, included measures of types of social–emotional instruction and children's social skills and challenging behaviors. Correlational analyses revealed statistically significant relationships between two types of social–emotional teaching practices and reductions in preschoolers' challenging behavior for children in the intervention group, teaching children behavior expectations and teaching problem-solving. No other statistically significant associations were identified. The results are discussed in relation to possible implications for social–emotional instruction for preschool children already exhibiting challenging behavior and the value of further research focused on preventing challenging behavior in young children.