Blended Learning has gained significant attention for its potential to enhance student achievement. However, its implementation in secondary education, especially in developing countries like Ethiopia, faces barriers such as infrastructure limitations, pedagogical challenges, and socio-economic factors. This study aimed to evaluate the effect of a blended learning model on the mathematics achievement of tenth-grade students in Ethiopia. A quasi-experimental design with non-equivalent groups, using both pre- and post-tests was employed. Two pre-existing classes from different schools were assigned to either the experimental group (n = 38) or the comparison group (n = 45) using lottery method. Prior academic records showed similar levels of mathematics achievement between the groups. A pre-test, covering topics like relations and functions, polynomial functions, and exponential and logarithmic functions, confirmed group equivalence. The data were analyzed using one-way ANOVA, which revealed that students in the blended learning group outperformed those in the conventional teaching group. A significant mean difference of 19.999 was observed, with a substantial effect size, as indicated by paired sample t-tests. These results suggest that blended learning is more effective than conventional methods in improving mathematics achievement. The study advocates for incorporating blended learning into the mathematics curriculum and urges policymakers and educators to consider adopting this model to enhance educational outcomes.