1. Histological evidence suggests that the force deficit associated with eccentric contraction‐induced muscle injury is due to structural damage to contractile elements within the muscle fibre. Alternatively, the force deficit could be explained by an inability to activate the contractile proteins. It was the objective of this study to investigate the latter possibility. 2. Mouse soleus muscles were isolated, placed in an oxygenated Krebs‐Ringer buffer at 37 degrees C, and baseline measurements were made. The muscle then performed one of three contraction protocols: (1) twenty eccentric (n = 10 muscles); (2) ten eccentric (n = 12); or (3) twenty isometric (n = 10) contractions. At the end of the injury protocol, measurements were made during performance of a passive stretch, twitch and tetanus. Next, force was recorded during exposure of the muscle to buffer containing 50 mM caffeine. 3. Decrements in maximal isometric tetanic force (P0) observed for muscles in the twenty eccentric, ten eccentric, and twenty isometric contraction protocols were 42.6 +/‐ 4.2, 20.0 +/‐ 2.3 and 3.9 +/‐ 2.4%, respectively. However, the caffeine‐elicited forces in muscles from the three protocols were not different when corrected for initial differences in P0 (64.9 +/‐ 1.3, 64.2 +/‐ 2.1 and 68.9 +/‐ 2.5% of pre‐injury P0). The peak caffeine‐elicited force was 118.4 +/‐ 8.6% of post‐injury P0 for the muscles in the twenty eccentric contraction protocol, which was significantly different from that observed for the other protocols (71.8‐80.2% post‐injury P0). These findings indicate that the force deficit in this muscle injury model results from a failure of the excitation process at some step prior to calcium (Ca2+) release by the sarcoplasmic reticulum. 4. In an attempt to locate the site of failure, intracellular measurements were made in injured muscles to test whether injury to the sarcolemma might have resulted in a shift of the resting membrane potential of the muscle fibre. However, microelectrode measurements of resting membrane potential for muscles in the twenty eccentric contraction protocol (‐74.4 +/‐ 0.6 mV) were not different from muscles in the twenty isometric contraction protocol (‐73.4 +/‐ 1.0 mV). These data suggest that membrane resting conductances were normal and are compatible with the idea that the ability of the injured fibres to conduct action potentials was probably not impaired.