The possible appearance of myocardial lactate production during exercise under hypoxaemia, simulating an altitude of about 4500 metres above sea-level (masl) was investigated. Twelve healthy men were studied, after coronary sinus catheterization, during prolonged exercise breathing 12% O2 compared with men breathing air. Coronary sinus blood flow was measured by thermodilution. Exercise duration under each breathing condition was 30 min and the order normoxaemia/hypoxaemia was varied between subjects so as to compensate for any influence of a preceding exercise period on a subsequent one. Work load was adjusted so as to produce a heart rate (HR) of 130-140 beats min-1 during both hypoxaemia and normoxaemia. [14C]lactate was infused at a constant rate i.v. throughout the study to detect a possible myocardial lactate release simultaneously with a net uptake. Myocardial O2 uptake did not differ significantly between hypoxaemia and normoxaemia. The compensation for reduced blood oxygen content was achieved entirely by a greater coronary blood flow. Yet, the arterial-coronary sinus (a-cs) lactate difference was lower during hypoxaemia than normoxaemia and isotope data indicated that this was caused by a myocardial lactate release of approximately 90 mumol min-1 which was at hand during hypoxaemia but not normoxaemia, whether hypoxaemic exercise preceded or succeeded normoxaemic exercise. In conclusion, A 27% reduction in arterial oxygen saturation is almost compensated for by an increased coronary blood flow. However, during hypoxaemic exercise cardiac energy demand is to a smaller part, about 1%, covered by anaerobic metabolism.