An a-posteriori multicentre reliability study was conducted to compare the reliability and magnitude of the maximum power (P[subscript max]) and optimal velocity (V[subscript opt]) between the force-power-velocity relationships during the leg cycle-ergometer and bench press throw exercises. The force-power-velocity relationships were determined in two separate sessions considering the values of force, velocity, and power collected under multiple loads. The force-velocity relationship was also determined from the force and velocity values collected under only the two most distant loads. Reliability of all regression models was high for P[subscript max] and V[subscript opt] during the leg cycle-ergometer (CV[subscript range]: 1.73-2.98%) and for P[subscript max] during the bench press throw (CV[subscript range]: 4.47-4.73%), while the reliability was acceptable for V[subscript opt] during the bench press throw (CV[subscript range]: 6.52-7.34%). Significant differences between the force-power-velocity relationships were observed for both P[subscript max] and V[subscript opt] (P < 0.001). All regression models can be confidently used to determine P[subscript max] and V[subscript opt], but these should not be used interchangeably.