Introduction: The triceps surae muscles, including medial and lateral gastrocnemius (MG, LG) and soleus (SOL), play a major role in supporting walking activities. Up to now, the potential selective recruitment of the muscles in walking of different intensities has not been investigated quantitatively. Such knowledge and its in vivo assessment would enable precise monitoring of subject’s performance in therapy. In this study, we measured apparent diffusion coeffcient (ADC) in the muscles of healthy subjects after plantar flexion of different intensities, and hypothesized that the different muscles would be selectively recruited in the exercises. Methods: 8 subjects were recruited. In a 3T MRI, subject performed straight-leg plantar flexion with load of 15 lbs for 3 minutes (moderate intensity), and the same exercise with load of 30 lbs until exhaustion (high intensity). Before and after each exercise, DWI of the right calf was performed: b values 0, 400 s/mm2, TE 70 ms, TR 4000 ms, 12 directions. In post-processing, ADC (unit: 10−6 mm2/s) was averaged for MG, LG and SOL. Results: In MG, ADC of 1580±12.3 before exercise increased to 1910±18.8 (p<0.01) and 2010±13.5 (p<0.01) after the moderate-and the high-intensity exercise, respectively. In LG, ADC of 1600±11.8 before exercise increased to 1900±14.8 (p<0.01) after the moderate-intensity exercise, and stayed stable at 1940±27.8 (P=0.30) after the high-intensity one. In SOL, ADC remained stable after the moderate-intensity exercise (from 1660±80.2 to 1660±11.5, P=0.26), and significantly increased to 1790±25.1 (p<0.01) after the high-intensity exercise. Discussion: The findings with MG and LG agree with expectation that muscle is more water-diffused with increased exercise intensity, and the association diminishes as exercise intensity increases. The higher ADC in MG than in LG at high intensity may be attributed to how the two heads attach to the Achilles tendon differently. The more linear attachment of MG to the tendon could enable higher effciency in achieving high torque. Attaching to tibia and fibula, SOL is not supposed to support straight-leg plantar flexion, as verified in its response after the moderate-intensity exercise. Surprisingly, SOL’s ADC increased after the high-intensity exercise. Besides the increased cardiac output as a possible cause, anatomical features of SOL may enable it to truly assist the exercise. First, similar to MG, SOL attaches to the Achilles tendon linearly, making its torque transfer effcient. Second, with more slow-twitch fibers than gastrocnemius, SOL’s activation is more economical in prolonged exercise. In conclusion, this study provided evidence that the triceps surae muscles are selectively recruited in straight-leg plantar flexion of different intensities, presumably to achieve maximal metabolic effciency. This work was supported by National NSF of China (#82171924, JLZ). No conflicts of interest, financial or otherwise, are declared by the authors. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.