This exploratory study encompassed a collaboration to implement and evaluate the effectiveness of Banking Time, a classroom intervention designed to promote supportive teacher-child relationships. Banking Time is a set of one-on-one meetings between a teacher and a child consisting of child-led play and teacher facilitation techniques. The study explored changes in teacher-reported relationship quality, teacher-rated child behavioral outcomes, and observer-rated teacher-child interactions during two six-week intervention periods. The three study conditions were (1) intervention; (2) within-class control; and (3) wait-list control. The sample consisted of 29 Head Start teachers and 116 children. Teachers completed the Student Teacher Relationship Scale (STRS; Pianta, 2001) and the Teacher-Child Rating Scale (TCRS; Hightower et al, 1986) at pre- and post-test. Teachers and children also participated in pre- and post-test videotaped semi-structured interactions that were coded on six teacher, child, and dyadic ratings. Results indicate that overall, there were modest classroom gains associated with the use of Banking Time. Teachers consistently reported increased perceptions of closeness with children participating in the Banking Time intervention. Teacher reports of classroom behavior reflected increased frustration tolerance, task orientation, and competence and decreased conduct problems as a function of the Banking Time intervention. Teacher beliefs, as measured by the Modernity Scale (Schaefer & Edgerton, 1985), were associated with ratings of child behavior as well as teacher-child interactions. Implications for prevention science research are discussed. (author abstract)