Objective: The objective of this study was to determine whether computer-based CPR training is comparable to traditional classroom training. Design and Setting: This study was quantitative in design. Data was gathered from a standardized examination and skill performance evaluation which yielded numerical scores. Subjects: The subjects were 64 undergraduate freshmen who never had CPR training or performed CPR. The subjects were divided into two groups. Group 1 completed the National Safety Council Adult CPR training program via traditional classroom instruction, while Group 2 completed the online version. Measurements: After training, both groups completed the standardized knowledge examination and skill performance evaluation. Skill performance was evaluated by trained evaluators and Resusci Anne computerized CPR manikins. Results: On the standardized knowledge examination and skill performance evaluation, Group 2 scored lower than Group 1; however, no statistically significant difference between the groups existed. MANOVA indicated there was a significant difference in the quality of CPR compressions (location, rate, depth, and release), ventilation rate and volume. Conclusions: Computer-based CPR training may be as effective as traditional classroom CPR training in terms of knowledge outcomes. However, the computer-based CPR training method may not be as effective as traditional classroom-based training in terms of developing quality CPR performance. These results are critical, as quality of CPR effort performed on a cardiac arrest victim has a direct impact on effectiveness. (Contains 4 tables and 1 figure.)