Studies with Hyhrid II dummies have demonstrated that the chest acceleration is a poor indicator of chest injury potential and that a realistic evaluation of the effectivencss of restraint systerxs should be based on measurement of the chest compression by means of strain gauges on a dummy's ribs. The projected introduction of the Hybrid III dummy into crash testing along with the requirement to measure sternum deflection for injury assessment has brought about the need to evaluate deflection measuring technology used on the Hybrid III. Tests revealed that different load distributions, i.e. due to a diagonal shoulder belt or an airbag, did not produce the expected different chest deflection patterns. That is, the thorax experienced the same deflection pattern with both restraint systems. This appears to be the result of an excessively stiff sternum assembly. To be sure that the highly protective effect of airbags, which is evident from real world accident analysis, be reflected in laboratory tests with dummies, this study suggests that the thorax of the Hybrid III dummy must be improved with respect to its sensitivity to injury producing local forces.