Recent studies suggest that non-neural cells can generate neurons in mice and humans. There is no lack of excitement about these results: they challenge our view of cellular differentiation and the prospect of being able to generate neurons from, for example, blood is very attractive. But is it too good to be true?