The goal of the federally-funded Reading First program is to ensure that all students read well by the end of third grade (Georgia Department of Education, 2006). However, Reading First makes few (if any) provisions for writing in its required 135-minute reading block for literacy instruction. Is it possible to teach reading effectively to young children without involving them in writing? The purpose of this naturalistic study was to investigate how the Reading First framework affected the teaching of writing in primary classrooms in one elementary school that received Reading First funding for three years. Using a social constructivist theoretical lens, the researcher explored these issues in the context of a professional learning community--a voluntary teacher study group--focused on writing instruction. Guiding questions were (1) What are primary teachers' perceptions of the reading-writing connection for students in kindergarten through third grade? (2) How does the context of a school wide Reading First grant affect primary teachers' perceptions of the reading-writing connection for students in K-3? (3) In what ways does a voluntary teacher study group focused on the reading-writing connection influence primary teachers' perceptions of the reading-writing connection and their literacy instruction? Fifteen primary teachers participated in the study during a six-month period. Data sources included an open-ended questionnaire, three in-depth interviews with each participant, audiotapes and selective transcription from ten teacher study group sessions, field notes from observations in 12 of the 15 participants' classrooms, a final focus group interview, and a researcher's journal. Data were analyzed inductively using the constant comparative method (Glaser & Strauss, 1967). Trustworthiness and rigor were established through methods that ensure credibility, confirmability, dependability, and transferability (Lincoln & Guba, 1985). Findings revealed that the teachers viewed reading and writing as connected processes in literacy instruction. Although the Reading First parameters made them fearful of engaging children in writing during the 135-minute reading block, the teacher study group validated their beliefs and knowledge and empowered them to interweave limited writing activities across the curriculum. Overall, the Reading First requirements prevented teachers from involving children in extensive writing process instruction and writing workshop.