Background: Studies comparing the effect of dynamic and static visualization suggest a predominantly positive effect of dynamic visualization. However, the results of individual comparisons are highly heterogeneous. In this study, we assess whether dynamic visualization (3D models and animations) used in the experimental group has a stronger positive influence on the intrinsic motivation and learning outcomes of science students (Biology, Chemistry and Geology) than static visualization used in the control group, and whether selected variables (students' gender, age, educational level, learning domain, and teacher personality) significantly affect the results. Results: This study was conducted in 2019 with a sample of 565 students from Czech middle (aged 11-15 years) and high (aged 15-19 years) schools using the following research tools: Motivated Strategies for Learning Questionnaire, Intrinsic Motivation Inventory and knowledge tests. The results show that using 3D models and animations in the teaching process significantly increased the students' intrinsic motivation for learning natural sciences (more specifically, its components (1) interest, (2) effort to actively participate in the educational process, (3) perceived competence and (4) understanding the usefulness of the subject matter), with a mean Hedges' g = 0.38. In addition, students in the experimental group reached a significantly higher level of Chemistry knowledge than their peers in the control group. Furthermore, by moderator analysis, we identified three moderator variables, namely student age, instructional domain and teacher personality. These variables significantly affect intrinsic motivation in different ways. The strongest positive effect of dynamic visualizations was found among students aged 11-13, whereas the weakest positive effect was identified among students aged 14-16. Regarding instructional domain animations and 3D models, the strongest positive effect is found in Chemistry (g = 0.74) and Biology (g = 0.72), whereas the positive impact on Geology is significantly weaker (g = 0.45). Teacher personality was found to be a major moderator in student motivation, with significant differences (g = 0.40-1.24). Teachers' attitude towards modern technology plays an important role concerning this effect. Conclusions: Based on these findings, we conclude that 3D models and animations have a positive effect on students and that teachers should include these visual aids in their lessons. For this reason, teachers are encouraged to implement these dynamic visual aids in their lessons regardless of their beliefs, and to get an adequate support in the process of implementation if necessary.