This study examined the relationship among baseline program and teacher characteristics and subsequent implementation of Banking Time. Banking Time is a dyadic intervention intended to improve a teacher's interaction quality with a specific child. Banking Time implementation was examined in the current study using a sample of 59 teachers and preschool children displaying disruptive behaviors in the classroom ([approximately]three children per classroom). Predictors included preschool program type, teacher demographic characteristics (personal and professional), and teacher beliefs (self-efficacy, authoritarian beliefs, and negative attributions about child disruptive behavior). Multiple measures and methods (i.e., teacher report, consultant report, independent observations) were used to assess implementation. We created three implementation composite measures (dosage, quality, and generalized practice) that had high internal consistencies within each composite but were only modestly associated with one another, suggesting unique constructs of implementation. We found that type of preschool program was associated with dosage and quality. Aspects of teacher demographics related to all three implementation composites. Teacher beliefs predicted dosage and generalized practice. Results suggest that the factors that predict the implementation of Banking Time vary as a function of the type of implementation being assessed. (author abstract)