It is well established that local muscle tissue hypoxia is an important consequence and possibly a relevant adaptive signal of endurance exercise training in humans. It has been reasoned that it might be advantageous to increase this exercise stimulus by working in hypoxia. However, as long-term exposure to severe hypoxia has been shown to be detrimental to muscle tissue, experimental protocols were developed that expose subjects to hypoxia only for the duration of the exercise session and allow recovery in normoxia (live low-train high or hypoxic training). This overview reports data from 27 controlled studies using some implementation of hypoxic training paradigms. Hypoxia exposure varied between 2300 and 5700 m and training duration ranged from 10 days to 8 weeks. A similar number of studies was carried out on untrained and on trained subjects. Muscle structural, biochemical and molecular findings point to a specific role of hypoxia in endurance training. However, based on the available data on global estimates of performance capacity such as maximal oxygen uptake (VO2max) and maximal power output (Pmax), hypoxia as a supplement to training is not consistently found to be of advantage for performance at sea level. There is some evidence mainly from studies on untrained subjects for an advantage of hypoxic training for performance at altitude. Live low-train high may be considered when altitude acclimatization is not an option.