Background: The delivery of effective, evidence-based early reading instruction depends partially on teachers' knowledge about reading acquisition, development and pedagogy. Research shows that teachers often perform poorly on measures of reading-related knowledge, often cited as one explanation for a gap between reading research and classroom practice. Studies on whether teacher knowledge improves student learning by leading to higher quality reading instruction, however, have produced mixed results. It is theorised that teachers' beliefs also influence the enactment of knowledge in the classroom. Yet, in comparison with investigations into teachers' knowledge, less research has explored their beliefs, including self-efficacy expectancy beliefs. Finally, although conceptually both knowledge and beliefs shape teachers' instruction, researchers often have limited information about what actually occurs in the classroom. Methods: We examined the knowledge, self-efficacy expectancy beliefs, self-reported literacy practices and literacy lesson plans of 34 teachers of reading working in four neighbouring rural schools to identify both strengths and areas for growth as potential targets for professional learning efforts. Results: On average, teachers demonstrated low levels of knowledge, reported implementation of both evidence-based and nonevidence-based practices and planned lessons that included only some essential components of a comprehensive early reading programme, with limited evidence of differentiation. Teachers overwhelmingly reported strong self-efficacy expectancy beliefs for teaching beginning readers. Conclusions: Addressing the reading research-to-practice gap requires deeper understanding of not only teachers' knowledge but also their perceived self-efficacy and instructional practices. Such information is critical for designing and delivering targeted, effective professional development.