The purpose of this study is to examine the effect of professional learning communities (PLCs) on elementary school students’ performance in reading and mathematics using data from the Early Childhood Longitudinal Study, Kindergarten Cohort of 1998 (ECLS-K). This study also investigates whether PLCs have differential effects on student performance based on student characteristics such as socioeconomic status (SES), race, and whether they are academically at-risk and school characteristics such as school type, school size, minority enrollment, and percentage of students eligible for free or reduced-price lunch (FRPL). PLCs are seen as a promising way of remedying the traditionally isolated nature of teachers’ work by facilitating a network through which teachers can share expertise, receive support, and disseminate effective practices. The underlying theory is that by facilitating teachers’ access to a network of their peers, they will be able to improve their instruction, which will ultimately lead to improved student achievement. This study addresses the need for more empirical evidence on the impact of PLCs on student performance using a large, national dataset. Principal component analysis (PCA) was used to identify correlated PLC items from the ECLS-K teacher questionnaire. Hierarchical and cross-classified random effects modeling (HCM) was then used to analyze the impact of student-, teacher-, and organizational-level variables—including two PLC variables—on students’ reading and mathematics performance. The analysis found that teacher collaboration had a significant positive effect on growth in reading and math scores, while a positive school climate was associated with significantly higher initial reading scores. Rarely did either PLC variable show differential effects based on student- or school-level characteristics.