Although literacy skills needed to engage in the economy and public life have grown, the literacy skills of many adolescents remain low. One popular approach to improving student literacy is using school-based reading coaches; however, there is little empirical evidence regarding the nature of coaching and its effectiveness in changing teacher practice and/or its impact on student achievement, particularly at the secondary level. In 2006-2007, RAND sought to address this research gap by studying a statewide reading coach program in Florida that is situated within a broader state-led literacy policy, the Just Read, Florida! (JRF) initiative. Using a combination of quantitative and qualitative methods to examine the implementation and impact of Florida's reading coach program at the middle school level, and to investigate three primary questions: (1) How is the reading coach program being implemented by the state, districts, schools, and coaches? (2) What has been the impact of coaching on teachers' practice, students' achievement in reading and mathematics, and other outcomes? and (3) What features of models and practices for reading coaches are associated with better outcomes? Reported findings include: (1) Districts established similar policies and supports for coaches (2) Common concerns about recruiting and retaining high quality coaches emerged; (3) Coaches' quality, particularly their ability to support adult learners, is positively related to several outcomes and viewed by some as an area of potential weakness; (4) Coaches indicated a desire for specific kinds of professional development; (5) Day-to-day work of coaches takes many forms; (6) District and school administrators, coaches, and teachers identified common barriers constraining coaches' ability and provide instructional support to many teachers; (7) Most coaches view school and district administrators as key supports for their work; (8) Many teachers and principals reported that the coach had positive effects on them and their schools; (9) The evidence is mixed regarding the impact of coaching on achievement; (10) The frequency with which coaches review assessment data with teachers is associated with positive outcomes and few other coaching implementation features were associated with student achievement. Based on study findings, the following recommendations are presented: (1) Provide guidance to school administrators in how to identify high quality coach candidates; (2) Develop a pipeline of qualified candidates; (3) Consider offering incentives and support to attract high-quality coaches and retain them over time; (4) Continue professional development for coaches with some adjustments; (5) Encourage coaches to review assessment data with teachers; (6) Address barriers to enable coaches to work more with teachers, including more one-on-one work; (7) If the intent is for coaches to work with all teachers, address barriers to working across the content areas; (8) Continue to nurture school administrator support; and (9) Continue research on coaching. Although increasingly popular, coaching requires a significant investment of both financial and human resources. As the field gains more evidence on the effects of coaching on teachers, schools, and students, researchers can work to determine whether the benefits of this intervention are worth the cost when compared with other interventions. Three appendixes are included: (1) 2006-2007 Florida Reading Coach Position Description; (2) Data and Modeling in the Statewide Achievement Analysis; and (3) Modeling the Coaching Implementation-Student Achievement Link. (Contains 17 figures and 49 tables.)