Motor vehicle crashes remain a significant problem in the US and worldwide. Automatic emergency braking (AEB) is designed to mitigate the most common crash mode:​ rear-end striking crashes. However, assessing the efficacy of AEB in real-world crash scenarios is challenging given that avoided crashes are rarely documented except during naturalistic driving studies. Unfortunately, a large-scale naturalistic driving study involving AEB-equipped vehicles has yet to be conducted. In the absence of such data, AEB can be evaluated in real-world crash scenarios by retrospectively adding AEB to naturalistic crash data using counterfactual simulations. Previous counterfactual simulations have purported the potential benefit of AEB;​ however these studies often make simplified assumptions about vehicle dynamics. To this end, the current study aimed to conduct the most realistic AEB counterfactual simulations to date by using measured host and lead vehicle dynamics data and vehicle-specific AEB deceleration profiles as well as accounting for driver reaction and environmental conditions. The SHRP2 Naturalistic Driving Study was reviewed to identify rear-end striking crashes among teen (16-19 yrs), young adult (20-24 yrs), adult (35- 54 yrs), and older (70+ yrs) drivers. Forty rear-end striking crashes that had reliable radar data were identified to serve as a basis for counterfactual simulations. Real-world AEB deceleration profiles were taken from IIHS AEB test data. IIHS AEB tests were matched to SHRP2 vehicles by selecting the most recent IIHS AEB test of the same make and vehicle class. AEB onset for SHRP2 crashes was based on a brake threat number (BTN) algorithm. The BTN was adjusted to match IIHS measured AEB onsets using minimum RMSE. AEB curves were then adjusted to match the speed of the subject vehicle at AEB onset. AEB deceleration curves were also scaled based on road surface conditions. Driver reaction was accounted for by beginning the deceleration curve at the current driver- initiated braking level. Counterfactual simulations were conducted using MATLAB to determine if AEB would have prevented the rear-end striking crash. AEB was found to be very effective, preventing 80% of rear-end striking crashes;​ greater than previously reported. Half of all crashes that were not prevented by AEB occurred during poor weather conditions. This study provides the most realistic counterfactual evaluation of AEB to date, utilizing real- world crash dynamics, driver reaction, road surface conditions, and measured AEB deceleration pulses. These data suggest that AEB is very effective at preventing rear-end striking crashes.