The biofidelity of the injury criteria used by Federal Motor Vehicle Safety Standards (FMVSS) No. 218 were examined against biomechanically based injury metrics. The current FMVSS No. 218 for motorcycle helmets is a generalized safety standard based on acceleration peaks and dwell times. FMVSS 218 uses a rigid headform mounted with a unidirectional accelerometer for conducting helmeted drop tests at a specified speed. The biomechanical basis of FMVSS No. 218 is not known, however. This study builds on previous work which developed a technique for measuring the under-helmet pressure contour on the headform during a FMVSS No. 218 impact attenuation test. By measuring the pressure applied directly to the head, the need for including the helmet in finite element analysis was bypassed. The headform pressure data from 80 impact tests to the front, crown and side of a helmet were used in finite element model simulations to predict skull fracture. FMVSS No. 218 impact attenuation injury criteria were correlated against the calculated skull strain and the generalized Skull Fracture Correlate (Chan et al., 2007 ESV). The acceleration data from the drop tests were used in NHTSA’s SIMon model to predict brain injury. The FMVSS No. 218 injury criteria were correlated against SIMon’s injury metrics for diffuse axonal injury (DAI), brain contusion, and subdural hematoma. FMVSS No. 218 injury criteria were also correlated against HIC. It was found that peak headform acceleration was the best correlate for all injury metrics. Dwell times over 150g and 200g both had very poor correlation with injury metrics. This research has shown that peak head acceleration can be an acceptable injury metric for the FMVSS No. 218 test method. However, the current FMVSS No. 218 limit of 400g allows for a high probability of skull fracture and brain injury.