We trace the history of the neo-Kraepelinian revolution which led to the publication of DSM-III. It is frequently overlooked that beginning with DSM-III there is a decided conflict between what is useful for clinicians and researchers. The conflict has more to do with the built-in research agenda of the DSMs introduced by and following DSM-III than with the nature of either clinical practice or research. We propose some fruitful ways for the two to work together. This requires incorporating a phenomenological approach to diagnosis and clinical research. With the centennial for the publication of the first edition of Jaspers "General Psychopathology" being celebrated the same year as appearance of DSM-5, it is surprising that little attention has been directed to the obvious dearth of phenomenology in the DSMs since DSM-III in 1980, and discussion about the relevance of phenomenology to current conceptual problems and controversies facing DSM-5. We propose ways that a fine-grained phenomenology suggestively cuts across the DSM-5 clusters indicating that phenomenology with its limited set of intersecting experiential dimensions—when practiced rigorously with fine-grained attention to patient experience—may lead to different hypotheses and results than the current thinking supporting DSM-5. In doing so, it heals the current artificial schism between top-down and bottom-up approaches to diagnostic classification.