Pérez-Castilla, A, Jaric, S, Feriche, B, Padial, P, and García-Ramos, A. Evaluation of muscle mechanical capacities through the 2-load method: Optimization of the load selection. J Strength Cond Res 32(5): 1245-1253, 2018-Recent research has revealed that the force-velocity relationship obtained from the "2-load method" (i.e., functional movements tested against just 2 external loads) could be a feasible method for the selective assessment of muscle force, velocity, and power-producing capacities. The study investigated the reliability and concurrent validity of the outcomes of the 2-load method observed from (a) farther vs. closer data points (20-70% 1 repetition maximum [1RM], 30-60% 1RM, and 40-50% 1RM) and (b) force-biased (50-70% 1RM) vs. velocity-biased (20-40% 1RM) data points. Twenty-two men were tested on a ballistic bench press throw performed against 6 incremental loads ranging from 20 to 70% of the bench press 1RM. The 2-load methods were constructed based on pairs of individually selected external loads and compared with the outcome of the force-velocity regression method applied to all 6 loads. The reliability and validity of the force-velocity relationship parameters decreased with the proximity of the data points (40-50% 1RM < 30-60% 1RM < 20-70% 1RM). The velocity-biased and force-biased loads (i.e., lighter and heavier loads, respectively) revealed a similar but relatively moderate reliability and validity. Overall, the farthest pair of loads (i.e., 20% 1RM and 70% 1RM) revealed the highest reliability (CV = 5.5%, ICC = 0.89) and validity (r = 0.98) among all the 2-load methods evaluated. These results demonstrate that the 20-70% 1RM 2-load method could be a feasible approach for testing individual muscle mechanical capacities, whereas the observed outcomes could be most reliable and valid when obtained from the farthest pairs of applied loads.