In order to discover how teachers structure the school day, this study examined the relationship between the sequencing, number, and duration of Gump's (1969) "activity segments" and the transitions between them. Detailed notes were written by an observer in 96 hours of observations of four primary and four intermediate classrooms. The observer focused on the teacher or student teacher who was primarily involved in directing classroom operations. After the 32 classroom chronicles were collected, they were segmented according to Gump's procedures. Transitions between activities were coded depending on the number of old segments that were terminating (or continuing) and the number of new segments that were beginning at the start of a transition. Results indicated that only 19 of the 42 potential transition structures were used. While three transition structures were very common, there was considerable variation among teachers in the number of transition types used. This variation was related to whether a teacher or student teacher was directing activities and to the grade level of the students. (CB)