Traumatic brain injury (TBI) is a major cause of death and disability. Despite its importance in public health, there are presently no drugs to treat TBI. Many reasons underlie why drugs have failed clinical trials, one reason is that most drugs to treat TBI lose much of their efficacy before patients are first treated. This review discusses the importance of therapeutic time window; the time interval between TBI onset and the initiation of treatment. Therapeutic time window is complex, as brain injury is both acute and chronic, resulting in multiple drug targets that appear and disappear with differing kinetics. The speed and increasing complexity of TBI pathophysiology is a major reason why drugs lose efficacy as time to first dose increases. Recent Phase III clinical trials treated moderate to severe TBI patients within 4–8 h after injury, yet they turned away many potential patients who could not be treated within these time windows. Additionally, most head trauma is mild TBI. Unlike moderate to severe TBI, patients with mild TBI often delay treatment until their symptoms do not abate. Thus, drugs to treat moderate to severe TBI likely will need to retain high efficacy for up to 12 h after injury; drugs for mild TBI, however, will likely need even longer windows. Early pathological events following TBI progress with similar kinetics in humans and animal TBI models suggesting that preclinical testing of time windows assists the design of clinical trials. We reviewed preclinical studies of drugs first dosed later than 4 h after injury. This review showed that therapeutic time window can differ depending upon the animal TBI model and the outcome measure. We identify the few drugs (methamphetamine, melanocortin, minocycline plus N-acetylcysteine, and cycloserine) that demonstrated good therapeutic windows with multiple outcome measures. On the basis of their therapeutic window, these drugs appear to be excellent candidates for clinical trials. In addition to further testing of these drugs, we recommend that the assessment of therapeutic time window with multiple outcome measures becomes a standard component of preclinical drug testing.