Fatigue-mechanisms seem to be different depending on whether the muscle is allowed to shorten (isotonic contractions) or not (isometric contractions) during stimulation. Slow twitch soleus muscles from 12 weeks old male Wistar rats (370 ± 4g) were stimulated intermittently at 30Hz (1s on 1s off for 15min; 1ms pulses) in situ at 37°C and allowed to shorten after the force had reached 1/3 of maximum force (234 ± 4g). Resting baseline-tension showed a transient rise, reflecting a significant slowing of relaxation that peaked after 100sec, and returned to near preload values after 5min. During the first 5min the shortening declined gradually to about 30% of the start values. The last ten minutes elapsed with no changes in the recorded parameters. The maximal isometric force-production was reduced after 100sec, but had regained pre-protocol level after 15min. Both exercising (EX) and contralateral resting (REST) muscles were frozen and analyzed for metabolites after 100sec and after 15min of stimulation. CrP and ATP were significantly lower in EX after 100sec, but there were no differences compared to REST after 15min. A 'slack test' showed that unloaded shortening velocities were lowest after 100sec (155mm/s), and approached pretest values (180mm/s) after 15min (165mm/s), indicating transient changes in the contractile apparatus. Multiplexed proteomics showed that phosphorylation of myosin light chain 2 (MLC2) was reduced at 100sec and remained low. We conclude that isometric muscle dynamics and velocity of unloaded shortening seem to follow CrP and ATP levels, whereas the isotonic fatigue, seen as reduced isotonic shortening, follows the MLC2 phosphorylation level.