Purpose of review The intrinsic limitations of the current major diagnostic classifications have become increasingly apparent. Several constraints these diagnostic systems impose on clinical practice and research have led to a widespread perception that psychiatry must move beyond the nosologic model embodied in contemporary classifications. Thus, the present review covers the ongoing debate on key conceptual issues that may be relevant for the development of ICD-11 and DSM-V. Recent findings The concepts of validity and utility, as applied to diagnostic classifications, received closer attention recently and attempts were made to better define and contrast them. Against such background, issues like conciliating the multiple purposes of diagnosis, choosing taxonomic strategies and determining the boundaries between and within (subtypes and specifiers) diagnostic categories and thresholds can be properly examined. Recent findings from genetics and neuroscience have added complexity to the already intricate field of mental disorders, raising high hopes of breakthroughs in pathophysiology. This bet on pathophysiology inscribes itself in the perennial debate about the nature of mental disorders and their defining (fundamental) level, which of course has critical implications for the future of psychiatry. Summary Despite considerable scientific advances achieved recently, the challenges faced by psychiatry have only been increasing in complexity. In the present state of affairs, conceptual closure and reductionism possibly represent the most serious risks to be avoided. The review ends by suggesting that methodological pluralism is a necessary condition for building a better psychiatric nosology.