AbstractThe models of reading and learning that underlie TRAY are examined, and criteria for selecting texts for classroom use identified. The kinds of language activities characteristic of groups using TRAY are described, and 15 categories of language activity are identified. The results of an analysis of the language activities of three groups are presented. The kinds of learning that are encouraged by TRAY are considered. Pupil and teacher evaluations are found to be favourable. TRAY is concluded to be an example of a good, content‐free, problem‐solving language development prog