This paper reports on a case study focusing on the development of students' capacity to develop and assess arguments in the context of instruction in high school genetics. It is part of a wider project whose goals were: (1) the identification of the conditions for argument (and in general scientific reasoning) to occur in science classrooms; (2) the analysis of argument patterns used by students; and (3) the exploration of the degree of specificity or subject-matter dependence of these argument patterns. The methodology involved observation, video and audiotaping of students while working in groups to design and solve problems. Toulmin's argument pattern was used as a tool for the analysis of students' conversation and this was coded using a framework for epistemic operations. The different arguments constructed by students are discussed along with what could be viewed as the students' version of the pattern (claims and warrants) which was used the most. The epistemic operations with relation to consistency and the evidence of school culture are also discussed. Implications for the context required for argumentation and true science dialogue in the classroom are suggested. Contains 22 references. (Author/JRH)