The problem of secondary school students‘ lacking understanding of mathematics remains a major concern to teachers and government authorities in Ethiopia. The problem has been largely attributed to using ineffective teaching methods. Research suggests the potential of mathematical discourse as a method of teaching mathematics for enhancing students‘ understanding of mathematical concepts and procedures. However, it remains a challenging practice for many school mathematics teachers to create classroom instruction that promotes meaningful mathematical discourse. One reason mentioned in the literature is a lack of a viable model for discourse-based mathematics instruction that assists teachers in how they create and enact discourse-based mathematics lessons in their classes. The other reason is limited research that explored the impact that any attempt to design and implement discourse-based mathematics instruction has on students‘ learning outcomes; in particular, in an Ethiopian secondary and upper school context. This study attempted to address the problem by developing a model for discourse-based mathematics instruction and evaluating its effectiveness in enhancing eleventh-grade students‘ conceptual and procedural understanding of probability and statistics as compared to a traditional teaching approach. To answer its research questions, the study employed the concurrent mixed research design that combines quasi-experimental and descriptive case study designs. The quasi experimental design with the posttest-control group that involved two control groups and one experimental group investigated the effects of discourse-based instruction on eleventh-grade students‘ conceptual and procedural understanding of probability and statistics in a randomly selected public secondary and preparatory school. At the start of the experiment, baseline data were collected from the three groups. A one-way analysis of variance of the baseline data revealed that the three groups were comparable. After the intervention, using a researcher developed test instrument, posttest data were collected from the three groups. The Kruskal Wallis test analysis performed on the posttest data showed significant differences in conceptual understanding and procedural understanding across the three groups. Pairwise comparisons of the groups using Dunn‘s test as Post-hoc analysis revealed that students who underwent discourse-based instruction demonstrated a better understanding of probability and statistics as compared to those students who received traditional instruction. The calculation of effect size xii demonstrated that discourse-based interventions have practical significance for improving students‘ understanding of various topics of probability and statistics. The descriptive case study design was used to investigate student experiences of discourse-based instruction. Qualitative analysis of lesson observations and a sample of student-written responses and the quantitative analysis of survey data provided insights into how discourse-oriented instruction impacted students‘ learning of probability and statistics. The findings revealed that a discourse-based instructional approach can be employed in classroom teaching and professional development programs. The findings may motivate educators to examine students‘ textbooks for mathematical discourse considerations. A possible recommendation for future research would be conducting longitudinal studies to examine how discourse-based mathematics instruction affects the relations between conceptual and procedural knowledge in a specific domain of mathematics. Keywords: mathematical discourse, conceptual understanding, instructional approach, quasi