In most of the somatic disciplines of medicine there is no doubt about the importance of diagnosis for patient care, research, teaching and regulatory affairs. In psychiatry however, the status of diagnosis has been antiguous for long times. In this paper the reasons for this difference will be discussed in the light of the historical development of diagnosis and classification in the beginnings of modern psychiatry. Especially during the period of antipsychiatric reasoning it was even regarded as harmful and dangerous for patients for be diagnosed in a scientific terminology. As a consequence the research in the field of mental disorder was hindered and blurred, and in addition also the scientific basis of our discipline was questioned. The American Psychiatric Association published in 1952 the „Diagnostic and Statistical Manual of Mental Disorders” (DSM-I), which was the first official manual of mental disorders to contain a glossary of descriptions of the diagnostic categories. A major breakthrough was the publication of DSM-III in 1980, which was not only characterized by explicit criteria for inclusion and exclusion, but also by strictly defined algorithms and by systematic field trials during the development of the manual. The DSM-approach to diagnosis in psychiatry became of major importance for research and practice in our field. Later, ICD-9 was transformed into ICD-10 by WHO with a certain tendency to adopt the basic principles of the DSM-system. Contemporary psychiatry is completely ruled by these two schemes of operationalized diagnosis. Advantages and disadvantages of this approach to classification for patient care, research and administration as well as possible alternatives and future directions will be analyzed. Special attention is given to the question, whether psychopathology and phenomenology are still relevant as methodological elements in psychiatry, even in an era of major progress of neurobiology in the sense of natural sciences.