In the early stages of recovery after a traumatic brain injury, patients usually experience a period of posttraumatic amnesia (PTA) in which they are confused, amnesic for ongoing events, and frequently behaviourally agitated. Although a variety of procedures and instruments are available to measure the duration of PTA, their applications in clinical practice are far from ideal. This review paper describes retrospective and prospective measures of PTA and critically examines variations in item content and scoring procedures. In particular, methods currently available to measure the amnesia component of PTA are especially problematic. The limitations of PTA scales give rise to a number of difficulties that impact upon clinical practice. These include determining precisely when a patient has emerged from PTA, and distinguishing between patients in PTA and those with chronic amnesia. It is concluded that there is a need to revisit basic constructs comprising PTA, and develop an instrument with greater specificity. More focus should be placed on examining the confusional component of PTA, and validity of PTA tests would be improved by measuring attention and behaviour, in addition to orientation and memory.