The accuracy and repeatability of crash data recorded by Generation 1, 2 and 3 Toyota Event Data Recorders (EDRs) in low-severity events were previously studied. The Toyota airbag control modules (ACMs) were subject to haversine acceleration pulses, and then the sensitivity of ACM response to characteristics of the crash pulse, EDR generation, and vehicle model was assessed. Linear regression models were created to accurately predict the reference speed change with a known ACM-reported speed change, pulse duration, peak acceleration, and vehicle model. The objectives of the present study are (i) to determine if accuracy trends in low-severity crashes can be applied to mid-severity collisions and (ii) to confirm our hypothesis that the ACMs function similarly on a sled as they did in a car, by comparing the response of ACMs subject to haversine, vehicle-to-vehicle, and vehicle-to-barrier collision pulses. First, a 2005 and a 2006 Toyota Corolla were used for a series of inline frontal and rear-end crash tests. 11 vehicle-to-barrier tests with speed changes ranging from 5 to 15 km/h and 10 vehicle-to-vehicle crash tests with speed changes between 4 and 14 km/h were conducted. Then, sixteen Toyota ACMs were mounted on a linear sled. The ACMs underwent frontal and rear-end haversine pulses for speed changes between 9 and 16 km/h. The ACMs were then subjected to replications of collision pulses from the in-vehicle tests. New linear regression models will be developed using the EDR data and statistical significant parameters will be determined. This study extends the understanding of how Toyota EDRs behave in low and mid-severity collisions.