This article brings a new method of measuring school misbehaviour. The method is known as the time reconstruction method. The time that is reconstructed in our research was defined as the period of two classes (2 x 45’) and one break (approximately 100 minutes in total). The students were asked to report their behaviour from minute to minute during the course of the reconstructed time. Total sample comprised 69 participants, 42 seventh-grade primary school students (16 males and 26 females) and 27 eleventh-grade students (17 males and 10 females). Their raw data were classified into 23 misbehaviour categories, with talking during the class as the most frequent category. We conclude that this method can gain the most objective results and because of that can be used for various comparisons, for tracking different changes in school life, and for evaluation of various models of classroom discipline.