This study was conducted to investigate whether or not computer-based simulations had a greater impact on science achievement compared to traditional hands-on methods for middle school students in an on-level science course. The study also sought to determine if either method had an impact on retention as well as motivation. The participants in the study were 6th grade students attending a public middle school in suburban metro-Atlanta. A variety of statistical analyses were utilized to measure science achievement, retention, and motivation. Results indicated that there was no significant difference on science achievement between the traditional hands-on method and the computer-based simulation method. While the control group and the experimental group both had academic gains, the control group experienced a statistically significant difference in gains on the density concept. There was no significant difference for the greenhouse effect concept in academic gains. Results further indicated statistically significant correlations between self-efficacy and science learning value, self-efficacy and active learning strategies, self-efficacy and achievement goal, and self-efficacy and performance goal.