Kuruvilla et al emphasise a point that we have also made in our study of acute and transient psychoses (Reference Marneros and PillmannMarneros & Pillmann, 2004), namely the heterogeneity of ‘acute psychoses’ which are diagnosed only on the basis of acute presentation. Unfortunately, the ICD–10 diagnosis of acute and transient psychoses relies primarily on mode of onset and uses symptomatology only for sub-classification. As we have shown (Reference Marneros and PillmannMarneros & Pillmann, 2004), a ‘polymorphic’, rapidly changing psychotic picture seems to be characteristic (as earlier authors always suggested) of the core group of acute polymorphous psychoses with good prognosis. In contrast, schizophrenia-like symptoms in the absence of polymorphic symptoms indicate a higher probability of later transition to schizophrenia. We believe that samples containing many of these patients with acute schizophrenia-like psychosis tend to lack the generally found female preponderance and show high rates of transition to schizophrenia early in the course. This is true for the sample of Amin et al (Reference Amin, Singh and Brewin1999) and may also apply to that of Thangadurai et al (2006). Thus, although we agree with Kuruvilla et al about the heterogeneity of acute psychosis, we feel that further refinement of our diagnostic criteria might help better delineate the core group of acute polymorphous psychoses.