Despite being a relatively new concept, psychiatric comorbidity, i.e. the co-occurrence of two or more mental disorders, has become widespread in clinical practice and psychiatric research. In this article, we trace the origin of the concept of psychiatric comorbidity, discuss the conceptual literature and point to basic problems concerning inadequate definition of the concept, differential diagnostic issues, and reification of mental disorders. We illustrate how these problems may have consequences for diagnostic assessment in current clinical practice and psychiatric research. To address some of the problems related to psychiatric comorbidity, we discuss potential principles for assessing psychiatric comorbidity. Inspired by Feinstein's original concept of comorbidity in general medicine and his differential diagnostic principles, we emphasize the importance of independence of mental disorders when assessing psychiatric comorbidity. We suggest that knowledge of trait v. state conditions and of the multitudinous clinical manifestations beyond what is captured in the diagnostic manuals may be helpful for assessing the independence of mental disorders and thus psychiatric comorbidity. We further argue that a more hierarchical diagnostic system and explicit exclusionary rules could improve clinical practice and research by reducing informational complexity and combating unwarranted psychiatric comorbidity.