The mechanisms of fatigue-induced changes in evoked compound muscle action potential (M-waves) are not well documented. In rats, isolated fast-glycolytic (tibialis anterior, or TA) and slow-oxidative muscles (soleus, or SOL) were stimulated repetitively at a low (10-Hz) or high (80-Hz) rate. Decreased amplitude and prolonged duration of M-waves were only significant after high-frequency fatigue in TA and SOL muscles; that is, in the conditions in which an influx of calcium was measured. On the other hand, maximal force failure and maximal lactic acid increase in the bath medium occurred in TA muscle after high- and low-frequency fatigue trials. Postfatigue increase in potassium concentration occurred in all circumstances. Thus, M-wave alterations depend mainly on the stimulation paradigm and not on the muscle type, and cannot be used as a marker of changes in intracellular potassium or lactate ions. This amplifies understanding of the significance of M-wave changes in human exercise protocols.