BACKGROUND: After 30 years of consensus-derived diagnostic categories in mental health, it is time to head in new directions. Those categories placed great emphasis on enhanced reliability and the capacity to identify them via standardized checklists. Although this enhanced epidemiology and health services planning, it failed to link broad diagnostic groupings to underlying pathophysiology or specific treatment response. DISCUSSION: It is time to adopt new goals that prioritize the validation of clinical entities and foster alternative strategies to support those goals. The value of new dimensions (notably clinical staging), that are both clinically relevant and directly related to emerging developmental and neurobiological research, is proposed. A strong emphasis on 'reverse translation' (that is, working back from the clinic to the laboratory) underpins these novel approaches. However, it relies on using diagnostic groupings that already have strong evidence of links to specific risk factors or patterns of treatment response. SUMMARY: The strategies described abandon the historical divides between clinical neurology, psychiatry and psychology and adopt the promotion of pathways to illness models.