Research on persuasive writing by elementary children posits primarily a developmental perspective, claiming that elementary-age children can effectively argue through talk but not through writing. While this view is commonly held, this article presents counterevidence. Drawing on two cases of third and fourth grade children writing persuasive letters gathered during six-month naturalistic studies of literacy practices and social identities in contrastive communities (one urban, one suburban), these data challenge the developmental generalization by showing that children in these settings can write persuasively. Further, this work complicates understandings of children's persuasive writing by showing how assignments and local cultures shape children's writing. Evidence is developed through rich description of the case study settings and instructional tasks, a typology of the children's persuasive strategies, and a critical discourse analysis of the children's persuasive letters. This study suggests that children in both communities are capable of persuasive writing, although they enact different patterns of response, drawing on locally learned discourses. The settings, the hybridity of the persuasive letter as both argument and letter, and the children's "habitus" may account for some of the differences in how the children address the tasks through ranges of centeredness and agentive strategies. Differing patterns of response suggest new frames for viewing and fostering children's argumentative competence in a range of settings, including understandings of agency. The author encourages a research agenda that accounts for socially situated classroom and community practices, and argues for ongoing research and critique of the power and place of persuasive writing for children in a range of schools. (Contains 7 tables, 1 figure, and 3 endnotes.)