Motorcycle accidents lead to a high rate of traffic mortality and morbidity. While helmet development and mandatory wearing have reduced head injuries, little progress has been made regarding trunk protection. Wearable airbag devices represent a promising solution to prevent trunk injuries. Nevertheless, research investigations need to be performed to assess and optimise the efficiency of such devices. This work consisted in the analysis of motorcyclist trunk impact conditions involved in various crash configurations to provide critical information in order to evaluate and improve the performances of airbag devices. First, an epidemiological and an accidentological analysis of data collection related to 252 real accidents, focusing on victims admitted into the shock rooms of two French trauma centres were performed. The data obtained was combined with numerical multibody parametric investigations, allowing the reproduction of 240 accident situations. An original and representative analysis of motorcyclists' impact conditions was provided, weighting the numerical study output data according to the real accident database. The impacted regions of the human body, the impact velocity and the accident chronology obtained in this work made it possible to define critical information for airbag efficiency assessment: the zones and levels of protection, the impacted surfaces as well as the airbag intervention time and the duration of maintained inflation of the airbag.