Changes in excitation-contraction coupling have long been recognized as possible causes of the failure in function which occurs in fatigued muscle. High-frequency fatigue is characterized by an excessive loss of force at high frequencies of stimulation and rapid recovery when the frequency is reduced. Frequencies in excess of 50 Hz are rarely seen with voluntary activation of human muscle, and for this reason there has been some doubt as to whether high-frequency fatigue is a significant feature of normal activity. Recent experiments have shown that with 30 Hz stimulation there is a more rapid loss of force if the muscle is held isometric in a shortened position and the fatigue is rapidly reversed if the muscle is re-extended, even under ischaemic conditions. These findings are consistent with the accumulation of K+ in the t-tubules and interfibre spaces of the muscle. Low-frequency fatigue is characterized by a relative loss of force at low frequencies of stimulation and a slow recovery over the course of hours or even days. There is evidence from intracellular measurements that low-frequency fatigue is due to a reduction in Ca2+ release. However, there is a possibility that in the fatiguing exercise, the end sarcomeras of the fibre overextend and damage those in the middle section of the fibre. In this situation the active sarcomas would be working at a shorter length than predicted from the overall fibre length and the force-frequency curve will be shifted to the right. Measurements of the length-tension relationship of muscles damaged by stretching are consistent with this happening.