Diagnostic stability is the measure of the degree to which a diagnosis remains the same at subsequent assessments of the patient and constitutes a longitudinal validation of the original baseline diagnosis. Follow-up studies including evidence of diagnostic stability and diagnostic consistency over time have been proposed to test the validity of psychiatric diagnoses. Until this moment definitions for psychiatric diagnoses are based on expert opinion rather than on their biological basis, and the modest knowledge base regarding underlying etiologies has hindered the use of etiological factors in psychiatric classification systems. But it is assumed that the higher the diagnostic stability, the more likely is to reflect a consistent psychopathological or pathophysiological process. Being that the main clinical purpose of diagnosis, as a formulation, is to furnish the informational basis for planning and conducting clinical care, stability of a diagnosis gives a relevant base not only for prediction of the course and outcome of a disorder but also for effective planning and provision of treatment. The availability of longitudinal data, however, may cause significant fluctuations in diagnostic stability as changes in clinical presentation are seen. Thus, evolving longitudinal observations should lead to periodic updating of the comprehensive diagnostic formulation, and yet, despite the inherent problems derived from criteria based on cross-sectional observations, our diagnostic system relies on stable diagnoses. Accounting for the potentially harmful consequences of unsuitable treatment options or clinical interventions, the study of diagnostic stability remains an essential issue in psychiatry.