Randolph Nesse argues that evolutionary theory is the key element in elaborating a valid criterion demarcating the normal from the pathological in psychiatry (Nesse, 2001, 2009, 2015, 2017). By focusing on the application of Nesse’s criterion on the demarcation of normal low mood from pathological depression, I argue – contrary to Nesse’s claims – that evolutionary theory cannot generate a valid criterion from the differentiation of normal low mood states from pathological depression. Indeed, expression in conformity to evolved functions cannot constitute a sufficient condition for normality, as Nesse’s project should imply. Moreover, grounding normality in evolved functions presupposes a kind of fixity of our emotional states, which seems unwarranted in the light of a constantly changing environment. As a result of these limitations, I suggest that the relevant distinction in psychiatry should not be “normality” versus “abnormality”, but, rather, “that which requires intervention” versus “that which does not”.