We examined the effects of a text-based argument writing intervention, "Triple Q," on argument writing skills in a cluster-randomized trial where groups of middle schools were assigned to conditions within school-level SES blocks. The intervention comprised three 15-day units. Students read and discussed argument texts representing different positions on a policy-related question; examined the features and quality of arguments within and across texts; provided feedback on peers' arguments; and drafted and revised their own arguments. The study was conducted in 27 sixth- and seventh-grade classrooms (k = 12 intervention) in 13 schools with 494 students (n = 220 intervention) in the southeastern United States. Results from differences-in-differences analyses showed that intervention students improved in overall argument writing quality compared to the control condition on a far transfer measure (ES = 0.44). To further explore the impact on specific dimensions of argumentation, we examined binary outcomes of conceptually robust indicators of argument writing skill. Results indicated greater increases in intervention students' likelihood of incorporating some dimensions of argumentation, such as supporting reasons with evidence and citing sources of evidence with specificity.