Research on adult readers' word recognition skills is used in this paper to develop a general information processing model of reading. Stages of the model include feature extraction, interpretation, lexical access, working memory, and integration. Of those stages, particular attention is given to the units of interpretation, speech recoding and the nature of lexical indices, and the possible influence of syntactic and semantic context clues on word recognition. Although some of the details about the mechanisms underlying those stages can be described, more research must be done on how skilled readers use context before the model can be applied to beginning reading instruction. Discussion following presentation of the paper is included. (RL)