We are grateful to Bracha et al for their interesting comments regarding primary agoraphobia as a potential evolutionary adaptation. First, we should clarify that we did not propose an additional diagnostic category; rather, we proposed that agoraphobia itself should be a stand-alone diagnosis in DSM–V (as in ICD–10), like other phobias. Subdividing what has historically been called agoraphobia may be useful, but we are concerned that clinicians and researchers are adopting Klein's narrower conceptualisation of agoraphobia as simply fear of panic in typical agoraphobic situations (Reference KleinKlein, 1980), without considering the possibility that a broader conceptualisation may be useful. Epidemiologists are increasingly adopting the definition of agoraphobia as ‘fear of fear’ (e.g. Reference Grant, Hasin and StinsonGrant et al, 2006), rather than the broader fear of difficulty in escaping, etc., in characteristic situations. As noted in our article, what has historically been called agoraphobia is strongly but not exclusively associated with panic, and, when the association exists, agoraphobia is not always preceded by panic.

The concept of an evolutionary basis for the development of phobias is not new (Reference SeligmanSeligman, 1971; Reference MarksMarks, 1987). Nevertheless, the reasons why people with agoraphobia develop fear and avoidance of particular situations remain important. Bracha et al suggest that fear of open spaces is an evolutionary remnant of primates’ use of trees to escape from predators. However, although some people with agoraphobia are fearful of open spaces, the list of typical agoraphobic situations is broad (Reference MarksMarks, 1987). Thus, hypotheses with an evolutionary basis to explain agoraphobia will be expected to cover reasons why persons fear and avoid a variety of situations. Although it is difficult to ‘prove’ such hypotheses, we agree with Bracha et al that researchers can make falsifiable predictions that can continue to illuminate the field.

We agree that cognitive–behavioural techniques may be particularly important for persons whose agoraphobia is primary. However, many people with agoraphobia can benefit from such treatment, whether the syndrome is primary or secondary (Reference KleinKlein, 1980).