This study aimed to compare the agreement between three-dimensional motion capture and vertical ground reaction force (vGRF) in identifying the point of dumbbell (DB) release during a countermovement jump with accentuated eccentric loading (CMJ), and to examine the influence of the vGRF analysis method on the reliability and magnitude of CMJ variables. Twenty participants (10 male, 10 female) completed five maximal effort CMJ at 20% and 30% of body mass (CMJ and CMJ, respectively) using DBs. There was large variability between methods in both loading conditions, as indicated by the wide limits of agreement (CMJ = -0.22 to 0.07 s; CMJ = -0.29 to 0.14 s). Variables were calculated from the vGRF data, and compared between four methods (forward integration (FI), backward integration (BI), FI adjusted at bottom position (BP), FI adjusted at DB release point (DR)). Greater absolute reliability was observed for variables from DR (CV% ≤ 7.28) compared to BP (CV% ≤ 13.74), although relative reliability was superior following the BP method (ICC ≥ 0.781 vs ≥ 0.606, respectively). The vGRF method shows promise in pinpointing the DB release point when only force platforms are accessible, and a combination of FI and BI analyses is advised to understand CMJ dynamics.