To evaluate the effects of cooling between exercise sessions on intramuscular water movement and muscle performance, the lower extremities of nine untrained men were assigned to either a cooling protocol (20-min water immersion, 15 °C) or a noncooling protocol. Each subject performed two exercise sessions involving maximal concentric knee extension and flexion (three repetitions, 60°/s; followed by 50 repetitions, 180°/s). The peak torque at 60°/s and total work, mean power, and decrease rate of torque value at 180°/s were evaluated. Axial magnetic resonance diffusion-weighted images of the mid-thigh were obtained before and after each exercise session. Apparent diffusion coefficient (ADC) values for the quadriceps and hamstrings were calculated for evaluating intramuscular water movement. Both groups exhibited significantly increased ADC values for the quadriceps and hamstrings after each exercise session. These ADC values returned to the pre-exercise level after water immersion. No significant difference was observed in muscle performance from first exercise session to the next in either group, except for increased total work and mean power in knee flexion in the cooled group. Cooling intervention between exercise sessions decreased exercise-induced elevation of intramuscular water movement and had some beneficial effects on muscle endurance of knee flexors, but not knee extensors.