Identifying effective ways to improve the quality of staff in early childhood programs, especially those that serve children in low-income communities, is recognized as an urgent national priority (Rhodes & Huston, 2012). Although a popular response in this funding and policy climate is to simply “change the workforce,” there is not a cadre of highly qualified and well-trained early childhood educators waiting to step in. Given this reality, coaching is being increasingly used as a means to improve teaching quality on the job and has shown considerable promise as a way to strengthen teachers’ classroom practices and improve child outcomes. A key unanswered question is how much coaching is needed to change teachers’ behavior and produce positive results for children? Research to identify the optimal dosage, duration, and type of coaching interventions is vital to improving the quality of the early childhood workforce (Ramey et al, 2011; Lloyd & Modlin, 2012). This panel first reviews how coaching models have been implemented across numerous contexts, including type of preschool setting, content of the coaching sessions, and varied participant background. We then summarize the findings about coaching interventions and associated outcomes, for teachers and children. Finally, we present the results from several randomized control trials (RCTs) of preschool interventions that employed coaching as part of a multi-component intervention and discuss the lessons learned from these studies. The preschool interventions all aimed to improve teacher instruction and teacher-child interactions through training, which were then practiced and reinforced through job-embedded coaching. Because the dosage and delivery of coaching varied across these studies, we have an opportunity to consider their implications for policy and future coaching practices, particularly in terms of adapting the intervention to the preschool setting and teachers’ individual needs.