Predicting the injury risk in automotive collisions requires accurate knowledge of human tissues, more particularly their mechanical properties under dynamic loadings. The present methodology aims to determine the failure characteristics of planar soft tissues such as skin, hollow organs and large vessel walls. This consists of a dynamic tensile test, which implies high-testing velocities close to those in automotive collisions. To proceed, I-shaped tissue samples are subjected to dynamic tensile tests using a customised tensile device based on the drop test principle. Data acquisition has especially been adapted to heterogeneous and soft biological tissues given that standard measurement systems (considered to be global) have been completed with a non-contact and full-field strain measurement (considered to be local). This local measurement technique, called the Image Correlation Method (ICM) provides an accurate strain analysis by revealing strain concentrations and avoids damaging the tissue. The methodology has first been applied to human forehead skin and can be further expanded to other planar soft tissues. The failure characteristics for the skin in terms of ultimate stress are 3 MPa±1.5 MPa. The ultimate global longitudinal strains are equal to 9.5%±1.9% (Green–Lagrange strain), which contrasts with the ultimate local longitudinal strain values of 24.0%±5.3% (Green–Lagrange strain). This difference is a consequence of the tissue heterogeneity, clearly illustrated by the heterogeneous distribution of the local strain field. All data will assist in developing the tissue constitutive law that will be implemented in finite element models.