1. The initiation of exercise‐induced muscle injury is thought to be the result of high tensile stresses produced in the muscle during eccentric contractions. Materials science theory suggests that high tensile stresses could initiate the injury during the first eccentric contraction (normal stress theory) or after multiple eccentric contractions (materials fatigue). It was the objective of this study to investigate the two possibilities. 2. Rat soleus muscles (n = 66; 11 protocols with 6 muscles per protocol) were isolated, placed in an oxygenated Krebs‐Ringer buffer at 37 degrees C, and baseline measurements were made. The muscle then performed an injury protocol which consisted of between zero and ten eccentric contractions (muscle starting length = 0.90 soleus muscle length, L0; length change = 0.25 L0; velocity = 1.5 L0/s; peak force = 180% maximal isometric tetanic tension (P0); time between contractions = 4 min; total duration of the injury protocol = 40 min). At the end of the injury protocol, the muscle was incubated in buffer for 1 h; every 15 min, an isometric twitch and tetanus were performed and lactate dehydrogenase (LDH) release was measured. Total muscle [Ca2+] was measured at the end of the incubation. 3. Change‐point regression analysis indicates that at 0 min into the incubation, declines in P0, maximal rate of tension development (+dP/dt), maximal rate of relaxation (‐dP/dt), and muscle stiffness (dP/dx) became significantly greater after eight eccentric contractions (p < or = 0.05). No relation was found between the number of eccentric contractions performed and the LDH activity at 0 min into the incubation, although after 60 min of incubation, LDH activity in the buffer was linearly related to eccentric contraction number (p = 0.01). There was no relationship between total muscle [Ca2+] and eccentric contraction number. These findings support the materials fatigue hypothesis of exercise‐induced muscle injury.