This paper focuses on examining the effectiveness of three-dimensional (3D) computer models on student teachers' academic achievement, mental model construction, and spatial ability used in learning the "atomic models" topic in this study. The students were randomly assigned into two groups: the treatment group (TG) where 3D computer models were used and the control group (CG) where models were not used for teaching. The treatment group was instructed using 3D computer models, while the traditional learning process was adopted in the control group. With the help of SPSS software, the independent-groups t-test and one way MANOVA were performed between the control and treatment groups. Cohen's d and eta-squared values were calculated for the effects of the computer models. Through this study, it was seen that the instruction using 3D computer models contributed to students' achievement, mental model construction, and spatial ability more than the traditional instructional process.