Most common preventive eccentric-based exercises, such as Nordic hamstring do not include any hip flexion. So, the elongation stress reached is lower than during the late swing phase of sprinting. The aim of this study was to assess the evolution of hamstring architectural (fascicle length and pennation angle) and functional (concentric and eccentric optimum angles and concentric and eccentric peak torques) parameters following a 3-week eccentric resistance program performed at long (LML) vs. short muscle length (SML). Both groups performed eight sessions of 3-5 × 8 slow maximal eccentric knee extensions on an isokinetic dynamometer: the SML group at 0° and the LML group at 80° of hip flexion. Architectural parameters were measured using ultrasound imaging and functional parameters using the isokinetic dynamometer. The fascicle length increased by 4.9% (p < 0.01, medium effect size) in the SML and by 9.3% (p < 0.001, large effect size) in the LML group. The pennation angle did not change (p = 0.83) in the SML and tended to decrease by 0.7° (p = 0.09, small effect size) in the LML group. The concentric optimum angle tended to decrease by 8.8° (p = 0.09, medium effect size) in the SML and by 17.3° (p < 0.01, large effect size) in the LML group. The eccentric optimum angle did not change (p = 0.19, small effect size) in the SML and tended to decrease by 10.7° (p = 0.06, medium effect size) in the LML group. The concentric peak torque did not change in the SML (p = 0.37) and the LML (p = 0.23) groups, whereas eccentric peak torque increased by 12.9% (p < 0.01, small effect size) and 17.9% (p < 0.001, small effect size) in the SML and the LML group, respectively. No group-by-time interaction was found for any parameters. A correlation was found between the training-induced change in fascicle length and the change in concentric optimum angle (r = -0.57, p < 0.01). These results suggest that performing eccentric exercises lead to several architectural and functional adaptations. However, further investigations are required to confirm the hypothesis that performing eccentric exercises at LML may lead to greater adaptations than a similar training performed at SML.