Personality and psychopathology have generally been regarded as distinct aspects of human behaviour, largely studied by researchers from different disciplines. However, an established body of research shows a common structure for personality and psychopathology phenotypes. This evidence has led to significant changes in how psychiatric problems are conceptualized and studied, as well as new questions about the differences between personality traits and mental disorders. In this Perspective, we suggest that models that differentiate personality and psychopathology based on the structure or stability of individual differences depict models of risk for psychopathology at the population level, rather than frameworks for psychiatric diagnosis at the individual level. We propose that person-environment transactions, across different timescales, hold the key for differentiating personality and psychopathology, and thus for psychiatric diagnosis. This proposal points to the need for new approaches to studying personality, psychopathology and the differences between the two.