Abstract: Mathematical modelling has been widely recognized as a learner-centred instructional approach that promotes conceptual understanding and problem-solving in mathematics. However, limited empirical evidence exists on its effectiveness in improving student-teachers’ achievement in matrices in Teacher Training Colleges (TTCs) in Rwanda. This study investigated the effect of mathematical modelling instruction on Year Two Science and Mathematics Education student-teachers’ achievement in matrices in selected TTCs in Rwanda. A quasi-experimental pre-test–post-test control group design was employed. The experimental group was taught matrices using mathematical modelling activities grounded in real-life contexts, whereas the control group received conventional instruction over a three-week intervention period. Data were collected using a researcher-developed achievement test administered before and after the intervention. Descriptive statistics and independent- and paired-samples t-tests were used to analyse the data. The findings indicated no statistically significant difference between the groups at the pre-test stage, confirming comparable baseline achievement. Following the intervention, the experimental group achieved a significantly higher post-test mean score (𝑀 = 60.64, 𝑆𝐷 = 9.57) than the control group (𝑀 = 50.30, 𝑆𝐷 = 8.83),t(101) = −5.70, p < .001. The experimental group also demonstrated a significantly greater improvement between pre-test and post-test scores than the control group. The study concludes that mathematical modelling instruction significantly enhances student-teachers’ achievement in matrices. It is recommended that mathematical modelling be integrated into mathematics instruction and teacher education programs to strengthen conceptual understanding and improve learning outcomes.