This paper presents results of an analysis of primary-aged pupils' educational experiences over a 35 year period. Data drawn from a set of large-scale systematic observation studies, conducted in the UK between 1976 and 2012, are used to describe pupils' average classroom experiences at six points in time over this period. These data are then used as markers for comparing the experiences of a subset of pupils - those with special educational needs (SEN) - over the same period. Results for the average pupil show an increase over time in the proportion of time spent interacting with teachers and peers. In contrast, relative to these non-SEN pupils, those with SEN have experienced a more moderate increase in the proportion of time spent interacting with the teacher, and almost no change in the amount of time spent interacting with peers and in whole class teaching contexts. The increase in the number of teaching assistants in mainstream primary settings, employed and deployed to assist the learning and inclusion of pupils with SEN, is identified as a key observable influence on the difference between the classroom experiences of pupils with and without SEN over time. This paper additionally defends the use of systematic observation methods, and concludes that the broad, but stable, measures of activity and behaviour, plus the rigorous approach to data collection it provides, are necessary for painting objective, descriptive and retrospective pictures of classroom life that can elude other research techniques.