This study examined the quality of teacher attributions for child disruptive behavior using a new measure, the Preschool Teaching Attributions measure. A sample of 153 early childhood teachers and 432 children participated. All teachers completed the behavior attributions measure, as well as measures regarding demographics, beliefs, self-efficacy, child behavior, and the quality of the teacher-child relationship with selected children. Confirmatory factor analysis demonstrated that the hypothesized 2-factor model fit significantly better than a 1-factor model, with the 2 factors being Causal and Responsibility. The resulting Causal and Responsibility subscale scores had solid internal consistency as measured by Cronbach's alpha coefficients. Significant bivariate and partial correlations with teacher practices and beliefs provided preliminary support for the measure's construct validity. Practice or Policy: Findings from this study suggest the importance of including a measure of teacher attributions in studies that explore teachers' beliefs, practices, and relationships with children. (author abstract)