Based on our review of the available data, we conclude that panic attacks are common among adolescents, while both panic attacks and Panic Disorder appear to be present, but less frequent, in children. Furthermore, it is evident that both adolescents and children who report panic attacks describe the occurrence of cognitive symptoms, although with less frequency than physiological ones. Consistent with the cognitive model of panic, it seems that at least some youngsters are capable of experiencing the physiological symptoms of panic accompanied by the requisite catastrophic cognitions. However, a more complete understanding of the cognitive manifestation of panic attacks/disorder among children awaits further investigation. Future research should aim to explore the developmental progression in children's cognitive responses to specific panic symptomatology. Risk factors (e.g. anxiety sensitivity, depression) which may contribute to the likelihood of misinterpreting physiological sensations in a catastrophic manner throughout the course of development should also be assessed: Finally, we are in general agreement with Abelson and Alessi (1992) who argue that we must begin to ask ourselves how panic disorder may be manifested in children. That is, rather than assessing the frequency with which children experience symptoms of adult panic, we should explore what panic would look like in children. They propose that the study of panic in children would be facilitated by a reformulation of separation anxiety as a childhood expression of panic disorder. Although this reformulation makes intuitive sense and is appealing from a developmental perspective, we would insert a strong caveat. Although the research is yet to be conducted, it is probable that childhood separation anxiety is only one of many routes to panic disorder outcome. It is improbable that such direct and continuous pathways are present for the majority of children, adolescents and adults who experience panic disorder. More probably, the pathways are multiple, complex, and discontinuous (Robbins & Rutter, 1990). Much work remains to be done before we are able to ferret out the linkages between developmental processes and clinical outcomes for panic disorder in children and adolescents.