The term "phlebotomy" is still sometimes used for the taking of blood for laboratory analysis or transfusion.
It was in use around the time of Hippocrates and was reinforced by the ideas of Galen, after he discovered the veins and arteries were filled with blood not air as commonly believed. There were two key concepts, first was that blood was created and then used up, it did not circulate and so it could 'stagnate' in the extremities. Second was that humoral balance was the basis of illness or health, the four humours were blood, phlegm, black bile, and yellow bile (relating to the four Greek classical elements of earth, air, fire and water). Galen believed that blood was the dominant humour and the one in most need of control. In order to balance the humours, a physician would either remove 'excess' blood (plethora) from the patient or give them a emetic or diuretic. Galen created a complex system of how much blood should be removed based on the patient's age, constitution, the season, the weather and the place. Symptoms of plethora were believed to include fever, apoplexy and headache.
The practice was also known in Islamic medicine, probably passed to them by the Greeks. Together with cautery it was central to Arabic surgery, the key texts, Kitab al-Qanum and especially Al-Tasrif li-man 'ajaza 'an al-ta'lif both recommended it. It was also known in Ayurvedic medicine, described in the Susrata Samhita.
Even after the humoral system fell into disuse, the practice was continued by surgeons and barber-surgeons. It was used to 'treat' a wide range of diseases, becoming an standard treatment for almost every ailment. It was especially popular in the young USA, where Benjamin Rush (a signatory of the Declaration of Independence) saw the state of the arteries as the key to disease, recommending levels of blood-letting that were high, even for the time. George Washington was treated in this manner following a riding accident, the withdrawal of almost 4 lb of blood certainly contributing to his death.
A number of different methods were employed. The most common was phlebotomy or venesection (often called "breathing a vein"), in which blood was drawn from one or more of the larger external veins, such as those in the foearm or neck. In arteriotomy an artery was punctured, although generally only in the temples. In scarification the "superficial" vessels were attacked, often using a syringe or cupping-glass. Leeches could also be used. The withdrawal of so much blood as to induce syncope (fainting) was considered beneficial.
Despite William Harvey disproving the basis of the practice in 1628 and the introduction of scientific medicine, la méthode numérique, allowing Pierre Louis to demonstrate that phlebotomy was entirely ineffective in the treatment of pneumonia and various fevers in the 1800s, in 1840 a lecturer at the Royal College of Physicians would still state that "blood-letting is a remedy which, when judiciously employed, it is hardly possible to estimate too highly" and Louis was dogged by the sanguinary Broussais, who could recommend leeches fifty at a time.
The reason for the continued popularity of blood-letting (and purging) was that while anatomical knowledge, surgical and diagnostic skills increased tremendously in Europe from the 17th century, the key to curing disease remained elusive and the underlying belief was that it was better to give any treatment than nothing at all, the psychological benefit to the patient outweighing the physiological downside. Even as blood-letting lost favour in the 19th century, there was nothing particularly to replace it. Outside of homeopathy, mercury, quinine, digitalis, amyl nitrate, and colchicum were the only substances widely used in medicine until late in the 19th century.
Phlebotomy still has its place in treatment of a few diseases like hemochromatosis.
See also: blood donation