The effectiveness of two professional development interventions, each designed to strengthen teacher-child interactions in preschool classrooms, was tested using a teacher-level randomized controlled trial. Georgia's Pre-K teachers (n = 486 in 336 schools/centers) were randomly selected from specified regions and randomly assigned to one of three conditions: 1) Making the Most of Classroom Interactions (MMCI), a cohort-model where small groups of teachers met for five days of instruction and support; 2) My Teaching Partner (MTP), in which teachers worked one-on-one with a coach using cycles of videotaped observations of teaching, review, and feedback; or 3) control. Each participating teacher received a Classroom Assessment Scoring System (CLASS; Pianta, La Paro, & Hamre, 2008) visit before and after the intervention from a trained, independent, blinded observer. Posttest scores were estimated as a linear function of condition and pretest score, using 2-level hierarchical linear models (HLMs). Findings indicated MMCI resulted in significantly higher posttest scores on Emotional Support and Instructional Support, and marginally higher posttest scores on Classroom Organization, as compared to controls. MTP resulted in significantly higher scores on Emotional Support. Findings are discussed in terms of implications for large-scale interventions to improve teacher-child interactions in early childhood programs. (author abstract)