To the Editor: I applaud Dr. Marshall et al. for their analysis of the validity and utility of the acute stress disorder symptom criteria and diagnosis; however, I do not draw the same conclusions from the findings reported to date. The fact that acute stress disorder falls short in predicting PTSD does not necessarily undermine its validity; many trauma victims show natural recovery over time (1), and some sufferers experience a delayed onset of symptoms. Moreover, the difficulty in differentiating normative and pathological posttraumatic reactions in the immediate aftermath of an event highlights the failure to identify the elements of the process that underlie the pathology, elements that apparently are not fully captured by the symptoms of either acute stress disorder or PTSD. In addition, findings regarding the predictive power of a variety of peritraumatic reactions (2), including dissociation, and reports of the ubiquity of dissociative symptoms in posttraumatic conditions (Butler et al., 1996) indicate that broadening our conceptions, rather than limiting them, may be most useful. Indeed, subtyping reactions (into, for example, types that principally involve dissociative versus hyperarousal or anxiety symptoms) may have utility. Individual differences, event characteristics, and features of the recovery environment (3) may also differentiate symptom profiles and courses. For example, individual differences in the facility or propensity to dissociate may represent a diathesis for the development of longer-term dissociative conditions, including PTSD, under conditions of extreme stress (Butler et al., 1996). The authors’ assertion that the two diagnoses cleave essentially continuous clinical phenomena actually begs the question.