To the Editor: We wish to express a number of concerns about the methods used in the study of mild traumatic brain injury by Hoge et al. (Jan. 31 issue).1 We learned through firsthand experi ence in the combat zone that immediately after a concussion, the soldier’s reported account is most accurate during the initial 24 hours after the blast incident.2 We therefore would be con cerned about recall bias introduced 3 to 4 months after a return from deployment. In addition, unlike perforation of the tympan ic membrane, a transient neurologic event (e.g., a “splitsecond” loss of consciousness) after a blast incident is not a sentinel finding of direct blast overpressure.2 Rather, it is a combination of at least three biodynamic variables: bodily displacement, directblast overpressure (involving intrinsic neuraltissue strains), and cardiopulmo nary insufficiency.3-5 The adverse effects of cardiopulmonary induced transient neurologic events (e.g., vasode pressor presyncope) are neither necessarily per manent nor cumulative, as a result of biologic restorative processes. Thus, we are concerned that the assessment methods used do not properly reflect the dimen sions of mild traumatic brain injury. Further more, we believe that the diagnosis of mild trau matic brain injury can be made only over time.