Amnesic effects of anesthesia were first reported two hundred years ago, but the term postoperative cognitive dysfunction (POCD) has appeared only recently, covering a larger range of neuropsychological modifications resulting from surgical intervention. The clinical description of POCD is highly variable, ranging from concentration impairment to delirium. Significant short-term POCD is common in elderly persons, and can persist several months, varying both in time and intensity and affecting the full-range of cognitive functions (visual and auditory attention, primary and secondary memory, implicit memory, and visuospatial functioning). Incidence rates vary widely according to surgery type but also between studies for a given surgical procedure, as a result of methodological difficulties and limitations. Variability is largely attributable to the absence of a standardized POCD definition, the heterogeneity of procedures to measure cognitive deficits and the methods used for statistical analyses, but is also related to the disparity in targeted populations. The wide variation in study design and target populations precludes the application of formal meta-analysis procedures. We review the definition, epidemiology, etiology, pathophysiology and the clinical and public health implications of POCD. The effects of anesthetics are described in relation to ageing-related physiological changes. It is concluded that the complex interaction of etiological factors makes it difficult to determine at this point of time to what extent POCD may be attributed specifically to anesthetic agents.