Cardiopulmonary resuscitation (CPR) training and its quality are critical in improving the survival rate of cardiac arrest. This randomized controlled study investigated the efficacy of a newly developed CPR training program for the public in a Taiwanese setting. A total of 832 adults were randomized to either a traditional or blended (18-minute e-learning plus 30-minute hands-on) compression-only CPR training program. The primary outcome was compression depth. Secondary outcomes included CPR knowledge test, practical test, quality of CPR performance, and skill retention. The mean compression depth was 5.21 cm and 5.24 cm in the blended and traditional groups, respectively. The mean difference in compression depth between groups was -0.04 (95% confidence interval -0.13 to infinity), demonstrating that the blended CPR training program was non-inferior to the traditional CPR training program in compression depth after initial training. Secondary outcome results were comparable between groups. Although the mean compression depth and rate were guideline-compliant, only half of the compressions were delivered with adequate depth and rate in both groups. CPR knowledge and skill retained similarly in both groups at 6 and 12 months after training. The blended CPR training program was non-inferior to the traditional CPR training program. However, there is still room for improvement in optimizing initial skill performance as well as skill retention. Clinical Trial Registration: NCT03586752; www.clinicaltrial.gov.