On reading the title of Dr Nair's letter, we were surprised that he appeared to suggest that there was no neuroimaging distinction between neurological and psychiatric disorders in our investigation. ^{Reference Crossley, Scott, Ellison-Wright and Mechelli1} After all, a direct comparison had shown statistical differences in several functional networks, as well as higher degree of similarity within each class than between the two classes. On reading the rest of the letter, however, it became apparent that the author's conclusion had little to do with the rigour of our methodology or the strength of our results. Rather, it was based on a more philosophical view that patients are not best served by the current distinction between the two classes.

We were puzzled by Dr Nair's suggestion that the observation of different brain structures for neurological and psychiatric disorders does not suggest segregation, and that a single dissociation, in which one class affects the brain and the other does not, would have provided greater evidence of segregation. First, this suggestion is methodologically difficult to sustain, since a double dissociation provides greater evidence of segregation than a single dissociation. ^{Reference McCloskey and Rapp2} Second, there is now compelling evidence that both neurological and psychiatric illnesses are disorders of the brain, and it would be misconceived to expect neuroimaging alterations for one class of disorders but not the other.

Dr Nair's interpretation seems to be based on the premise that patients are not best served by the current classification – the empirical data, however, suggest that there is a neuroimaging distinction between neurological and psychiatric disorders. A more nuanced approach is to recognise, as we do in the manuscript, that ‘neuroimaging evidence does not necessarily mean that the existing distinction is useful from a clinical perspective’. In other words, neuroimaging evidence should be considered one of several factors informing this debate; negating such evidence, in contrast, will only cloud the debate.

In conclusion, the clinical rationale for combining neurological and psychiatric disorders into a single category, as well as the opposite view that this would be detrimental to patients, have been discussed extensively elsewhere. ^{Reference White, Rickards and Zeman3–Reference Bailey, Burn, Craddock, Mynors-Wallis and Tyrer4} The aim of our manuscript was to help refine this debate by providing an alternative perspective based on current neuroimaging evidence. We believe that the neuroscientific perspective cannot be discarded if we are to develop integrated mind–brain models of disease that can be translated into clinical practice. Only this will enable psychiatry to become a ‘brain-based medicine of the mind’. ^{Reference Bullmore, Fletcher and Jones5}