This study examines the effects of a teaching method using 3D virtual reality simulations on achievement and attitude toward science. An experiment was conducted with fifth-grade students (N = 41) to examine the effects of 3D simulations, designed to support inquiry-based science curriculum. An ANOVA analysis revealed that the 3D group scored significantly higher on the achievement test (F = 7.03, p < .05) than the control group using traditional 2D visuals. A positive change of attitude toward science was identified in both 3D and 2D groups, but it was not statistically significant. A robust regression analysis revealed that neither gender nor ethnicity had any significant effect on achievement tests or attitude. However, prior attitude had a significant determining effect for later attitude in both groups.