Within the last 10 years, new tools for assisting in the teaching and learning of academic skills and content within the context of science have arisen. These new tools include multiple types of computer software and hardware to include (video) games. The purpose of this study was to examine and compare the effect of computer learning games in the form of three-dimensional serious educational games, two-dimensional online laboratories, and traditional lecture-based instruction in the context of student content learning in science. In particular, this study examines the impact of dimensionality, or the ability to move along the X-, Y-, and Z-axis in the games. Study subjects (N = 551) were randomly selected using a stratified sampling technique. Independent strata subsamples were developed based upon the conditions of serious educational games, online laboratories, and lecture. The study also computationally models a potential mechanism of action and compares two- and three-dimensional learning environments. F test results suggest a significant difference for the main effect of condition across the factor of content gain score with large effect. Overall, comparisons using computational models suggest that three-dimensional serious educational games increase the level of success in learning as measured with content examinations through greater recruitment and attributional retraining of cognitive systems. The study supports assertions in the literature that the use of games in higher dimensions (i.e., three-dimensional versus two-dimensional) helps to increase student understanding of science concepts.