BackgroundPrevious studies used curriculum-based measurement (CBM) maze scores as an indicator of the reading comprehension level of secondary school students with and without special educational needs in multiple grades, pinpointing a high influence of both student- and context-related variables. However, studies on cumulative influence are necessary for better understanding of data-based decision-making.MethodsWe examined a sample of 1066 secondary school students using four linear mixed-effect models: How much variance in maze scores exists between multiple student characteristics (i.e., gender, immigration background, learning disability and developmental language disorder) and context variables (i.e., classroom, grade and school type) across Grades 5-8?ResultsThe intra-class correlation (ICC) results show that the influence by the context-related variable classroom (ICC = .094) is almost as large as by the variable grade level (ICC = .126). School type (i.e., inclusive school vs. special school) has the least influence (ICC = .02). In addition, the effects of student-related variables explain only a small proportion of the variance (marginal R2 = .114).ConclusionsMaze scores can be used as a screening instrument for students with multiple characteristics across grades; they also show that it makes no difference which type of school students attend. As teachers and further classroom-related variables have almost as much influence as grade level, we discuss that teachers can minimise classroom effects by using maze scores as a formative approach. What is already known about this topicCBM maze scores can be used as indicator for reading level as a screening and formative assessment.Individual differences at student and context levels uniquely influence reading.Teachers use maze scores for decision-making at both an individual student level and a classroom level to prevent reading difficulties.What this paper addsStudent variables only are not sufficient to explain variance in maze scores.The impact on maze score by classroom membership is almost as large as by grade level.The type of secondary school students attend makes no differences in terms of maze scores.Implications for theory, policy or practiceTo explain variance in maze scores, a joint consideration of the influence by student- and context-related variables is necessary.When planning reading instruction, particularly, classroom influences need to be considered.