A study was conducted to evaluate the impact of predictable stories given to children from low-income families over a two-year period. Phase one subjects were all the children (N=52) from four Headstart classes; phase two subjects were all the children who had participated in phase 1 and who entered kindergarten after their last year in Headstart. The children were assigned to experimental and alternative groups; the remaining children were placed in the control group. The experimental group received six predictable stories during Headstart and six during kindergarten to read at home. The control group received pictures of six familiar children's stories during Headstart and six sets of workbook activities during kindergarten. Both groups received six lessons to learn about the materials during the Headstart year. In May of the Headstart and kindergarten years, children were given spelling and reading subtests and were asked to read old and new stories. Parents completed questionnaires regarding their child's use of the materials that had been sent to their homes and their child's interest in these materials and their knowledge about reading and writing. Results indicated that the experimental group scored significantly higher on story reading, word reading, and spelling. Teachers reported that a significantly higher number of children from the control group might have problems with reading when they began first grade. Parents of the experimental group rated their children significantly higher on questions concerning their child's interest in and knowledge about reading and writing than did the parents of the control group. Findings suggest that predictable stories encourage children to behave like readers at home and that this at-home activity influences early reading development. (Eight tables of data including a detailed analysis of responses to the parent questionnaire are included and 38 references are attached.) (Author/MG)