The purpose of this study was to compare three different methods for the estimation of vertical jump height and to reveal errors in maximum height of the centre of gravity (CG) that emerged. Additionally, the problem of initial values, which are difficult to estimate for squat jumps and drop jumps, was addressed by suggesting a new method for the calculation of CG height of squat and drop jumps. To evaluate and compare the methods used to examine the vertical jump performance, 15 men performed four maximal vertical jumps of three different types: counter movement jump (CMJ), squat jump (SJ) and drop jump (DJ). Kinematic data and the vertical ground reaction force were recorded. Three methods were used for the evaluation of the jumps: flight time method (FT), impulse method based on the energy conservation law (IMP) and the double numerical integration of the ground reaction force (INT). The maximum CG height calculated from three-dimensional (3D) kinematics using automatic infrared (IR) marker tracking (Vicon [Oxford, UK], Plug in Gait model, Davis et al., 1991) was used as reference (golden standard). The results of the 3D kinematic analysis and the double integration of the vertical reaction force did not show significant differences. The jump heights determined by IMP and by FT were too low in relation to the real jump height due to the absence of the CG height at take off. Besides this, the flight time can be manipulated by the jumper, meaning that the flight time method is not suitable for CG height estimation of vertical jumps in scientific studies.