Objective: Traditional cardiopulmonary resuscitation (CPR) training programs do not target older adults who are most likely to witness private-residence cardiac arrests and do not reliably result in a bystander who is likely to perform CPR in the event of an arrest. This study was performed to compare targeted CPR training programs for older adults (older than 50 years) that 1) increase numbers of CPR-trained bystanders of private-residence cardiac arrest or 2) increase the percentage of trained bystanders of private-residence cardiac arrest who perform CPR. A simultaneous outcome was to estimate the minimal significant survival benefit associated with each of the training programs. Methods: A probabilistic simulation model was developed in Fortran95 that incorporated key out-of-hospital cardiac arrest elements, including witnessed arrests, CPR-trained witness, CPR provision, and impact of CPR on ventricular fibrillation. Input data were derived from published or publicly available data, including a large prospective cohort study of outcomes in Oakland County, MI. Monte Carlo simulation (n= 10,000) and sensitivity analyses (n= 40) were used to assess median and the empiric 95% confidence intervals [CIs] for incremental survival with either intervention. Results: The baseline model, calibrated to the characteristics of the input-data community, established that, for private-residence cardiac arrests, 40.8% of cardiac arrest bystanders were trained in CPR; however, only 25.7% performed CPR. This yielded 4.81% survival (95% CI = 4.72 to 4.89). Modeling the impact on the baseline training level with increased CPR performance among trainees indicated that 75% of private-residence trained bystanders would need to perform CPR in order to reach a minimally significant improvement in survival (5.02%; 95% CI = 4.94 to 5.15). Similarly, targeted CPR training that would result in a significant survival benefit (to 5.01%; 95% CI = 4.93 to 5.09) would require that 70.8% of bystanders be trained. Conclusions: CPR training programs that focus on yielding 75% of trainees who perform CPR in the event of witnessing an arrest would have equivalent results to mass CPR training programs that result in 70% of bystanders being trained in CPR. However, the minimal survival benefit associated with these programs (around 0.2%) may prove either method costly with minimal effect.