The present study reports on the psychometric properties of the Early Childhood Classroom Observation Measure (ECCOM) in Finnish and Estonian first and third grade classrooms.The observation data were collected from 91 first grade teachers (32 from Finland, 59 from Estonia); and 70 third grade teachers (33 from Finland, 37 from Estonia). Teachers also selfrated variables, such as curriculum goals, their teaching experience and the classroom size.The results of confirmatory factor analysis (CFA) provided evidence which supported use of the three-factor model (Management, Climate, and Instruction) for each dimension that is, child-centred, teacher-directed and child-dominated, in both grades. The reliability of the dimensions and sub-scales was good, and some evidence was also found for criterion validity.The findings imply that ECCOM is thus a reliable and valid instrument for measuring teaching practices in primary school.Keywords: teaching practice, child-centred classroom, teacher-directed classroom, child- An increasing amount of empirical research has indicated that teaching practices in preschool, kindergarten and the early years of school can play a significant role in student achievement later on (e.g., Hamre & Pianta, 2001;Pressley et al., 2003). However, much of what we know about classroom practices in schools comes from surveys and one-on-one standardised assessments (e.g., Flowers & Hancock, 2003;Ireson, Blatchford, & Joscelyne, 1995;Kiuru et al., 2012). The problem with those measures is that they lean heavily on selfreports from teachers and/or students, which means that they have the tendency to be biased.Consequently, it has been suggested that there is a need to find more concrete indicators of classroom practices through more effective classroom observation (e.g., Hamre & Pianta, 2005;La Paro, Pianta, & Stuhlman, 2004). What needed are observational methods that take into account the instructional and organisational choices that teachers make to create the classroom climate (e.g., Pianta, La Paro, & Hamre, 2008;Stipek & Byler, 2004). One such observational method is the Early Childhood Classroom Observation Measure (ECCOM; Stipek & Byler, 2005). However, previous studies on ECCOM have some limitations. First, it has so far been almost exclusively used in preschool and kindergarten (e.g., Lerkkanen et al., 2012b;Rao, Sun, Zhou, & Zhang, 2012;Yen & Ispa, 2000). Although there are a few studies (Kikas, Peets, & Hodegs, 2014;Lerkkanen, Kikas, Pakarinen, Poikonen, & Nurmi, 2013;Perry, Donohue, & Weinstein, 2007) which have used it in the first grade classrooms, there are no studies which have investigated the reliability and validity of ECCOM later on in primary school grades. Second, the age that children start kindergarten and primary school vary from one country to another. For example, first-graders in Finland and Estonia are somewhat older than first-graders in U.S., where the instrument was developed. Third, to develop observational instruments information is always needed on the appli...