An extracorporeal cardiopulmonary resuscitation (eCPR) is considered as a therapy option for cardiovascular failure that is refractory to therapy. It can significantly improve the survival rate with favourable neurological results in highly selected patients. The initially defibrillatable heart rhythm and the short low-flow time < 60 minutes are of particular prognostic value. An essential prerequisite for deciding on eCPR is the existence of a reversible cause for cardiac arrest. Whether an eCPR directly at the emergency site (out-of-hospital variant) or in the clinic, e.g. in the cardiac catheterization laboratory (in-hospital variant) can be recommended must be clarified in further randomized-controlled, multicentre studies. Both variants have advantages and disadvantages. With the out-of-hospital eCPR, the "collapse-to-start-eCPR-time" can be significantly reduced under certain conditions. With the in-hospital eCPR external negative influences can be greatly minimized.