The present thesis tested the reliability and validity of the Early Childhood Classroom Observation Measure (ECCOM; Stipek & Byler, 2004, 2005), identified the profiles of teaching practices, i.e., teaching styles, and, finally, examined the predictive role of teaching styles in early primary students’ reading skills. Specifically, the thesis had four aims: first, to examine the psychometric properties of the ECCOM in Finnish and Estonian early primary school, i.e., grades 1 and 3; second, to identify the profiles of teaching practices, i.e., teaching styles, among Finnish and Estonian teachers by using a person-oriented approach; third, to compare the use of teaching styles between Finnish and Estonian teachers; and, finally, to investigate the predictive relationships between teaching styles and reading skills. The data stem from the First Step longitudinal study in Finland, and the Kindergarten-School Study and the Reading Study in Estonia. A total of 91 first-grade teachers and 70 third-grade teachers were observed using the ECCOM instrument. One thousand and twelve first-grade students and 958 third- grade students in the classrooms of these teachers were examined in terms of their reading skills, i.e., fluency and comprehension. A subgroup of 359 Finnish students reading skills was tested longitudinally from grade 1 to grade 3. The results of the present thesis firstly suggest that the ECCOM is a valid classroom observation tool in Finnish and Estonian early primary grades. Secondly, the thesis identified four latent profiles of teaching practices in the first grade: child-centred, teacher-directed, child- dominated, and mixture teaching style, i.e., a mixture of child-centred and teacher- directed practices. In addition, one extra profile was found in the third grade: extreme- child-centred style. Thirdly, Finnish teachers used less of the child-dominated style in grade 1 and more of the child-centred style, but less of the extreme-child-centred style in grade 3 than Estonian teachers. Finally, the child-centred style and mixture teaching style showed more beneficial influences on students’ reading skills development than other teaching styles. By contrast, the child-dominated style was found to have detrimental impacts on children’s reading skills development. The extreme-child-centred style did not, however, contribute positively to reading fluency in grade 3. Overall, the results suggest that the ECCOM is a valid and reliable measure of classroom practices in school environments outside the United States. Moreover, the use of the person- oriented approach can expand our understanding of teaching practices in important ways, and child-centred and teacher-directed practices can be used in conjunction to support children’s reading skills development in early school years.