This work sought to obtain experimental evidence to corroborate cross-sectional patterns of development in argument skills and to evaluate the effectiveness of an intervention designed to foster development of these skills in academically at-risk 13- to 14-year-olds. Students participated in 16 sessions of a collaborative, goal-based activity providing dense exercise of argumentive thinking. One condition included peer dialogues; another did not. The former was the more effective, although both groups progressed. Participants showed increased frequency of usage of powerful argumentive discourse strategies, such as counterargument, and decreased frequency of less effective strategies. Quality of individual arguments (for or against a claim) also improved, supporting the existence of a close relation between these two kinds of argument skills.