The medical concept of 'fever' has undergone profound changes throughout the centuries. Galen of Pergamon considered fever as a systemic disease in itself, and it was only between 17th and 18th century that Hermann Boerhaave provided a more careful evaluation of the clinical phenomena related to fever. Apart from incorrect theories, a major obstacle to the development of a rational study of fever has been the lack of appropriate instruments of measurement; in effect, the clinical thermometer was not diffusely used in everyday medical practice until the mid 19th century. During this same period Ignaz Semmelweiss postulated that the pathological-anatomical changes recorded in women who had died because of puerperal fever represented a pathological reality clinically suggested by a whole cohort of symptoms and signs, among them fever. Even if enormous progress has been made in the 20th century with regard to fever, which is currently considered a clinical sign of many different diseases, its etiologic assessment remains a challenge. In fact, in 1961 the clinical picture of 'Fever of Unknown Origin' was officially defined. Since such diagnostic labelling is in effect a cover for our inability to discover the real causes of fever, in this case, paradoxically, fever goes back to being the whole pathological picture, just as it was retained to be many centuries ago.