In response to the assertion that reading is a fixed set of processes that can be identified and studied independently of context, a study was initiated based on the following assumptions: that different types of reading (prose comprehension and text search) can be identified and measured, that reading competencies are highly associated with uses for reading, and that reading proficiency can be predicted by the frequency of the use of each type. The objectives of this study, which is in process, are to demonstrate a new methodology for assessing reading skills, to develop a survey instrument to relate materials read to the uses to which they are put, and to construct a reading competencies assessment instrument. The sample population, 25 employees of Hewlett-Packard, have been observed at work. Sample tasks are currently being compiled. Findings thus far include the following: (1) independent sets of reading competencies are developed in relation to particular uses; (2) time spent engaged in reading for a particular use is more important in predicting performance than any other background or performance variable; (3) reading for a particular use does not substitute for other types of reading; and (4) different types of materials require different reading skills. Further research is necessary to replicate these findings. (Appendixes contain the text search processing model, a chart depicting the activities time line, an outline of the briefing for Hewlett-Packard, and a copy of the initial sampling plan.) (LLZ)