A year-long study of the writing development of 27 first through third graders in an English/Spanish bilingual program was conducted during the 1980-81 school year. Samples of the children's writing were collected at four intervals, coded for computer tallying, and analyzed in terms of code-switching, spelling, punctuation and segmentation, structural features, stylistic devices, and content. Additionally, the context in which the writing developed was evaluated by classroom observations, teacher interviews, review of familial backgrounds, and a survey of the community language situation. Myths about bilingual language proficiency, biliteracy, bilingual education, teaching writing, and learning to write are all countered by evidence presented in this study. In a discussion of implications, the concept of a whole language approach to writing instruction is supported, in which authentic and functional texts are offered to and produced by children. Examples of the children's writing with appropriate translations are given along with various tables. Informal follow-up information is presented in three epilogues dealing with changes in the researcher's commitment to the study's original writing theories, the writing of some students a year after the study; and a chronological outline of the demise of the bilingual program used in the study. Appendices list interview questions used for teachers and aides and categories for coding the writing data. This book contains 134 references. (ALL)