This paper reports a technique for classroom observation of children's activities considered relevant to the acquisition of reading. It is part of a larger study conducted by the University of Bristol called the "Children Learning to Read" Project, an attempt to study the learning experiences of individual children as they occur in a variety of schools. The method is to explore the interaction between oral language ability, relevant preschool experience and initial motivation with respect to reading; range and types of texts and insturctional materials; and relative emphasis on timing in strategies of teachers concerned in the initial stages of reading instruction. The sample is 20 infant-school children. Observations take the form of a running record of an individual child's activities during a complete morning once every month over a two-year period. The record is designed to yield an account of the different activities engaged in and the amount of time spent on each; and a detailed description of significant aspects of these activities. A coding system is developed for recording relevant information. Results are in the form of prose records and fully-coded time distribution records. Appendixed is a portion of one observation to illustrate the precise form. (Author/MS)