In this paper, an experimental analysis and validation of a minimum time-jerk trajectory planning algorithm is presented. The technique considers both the execution time and the integral of the squared jerk along the whole trajectory, so as to take into account the need for fast execution and the need for a smooth trajectory, by adjusting the values of two weights. The experimental tests have been carried out by using an accelerometer mounted on a Cartesian robot. The algorithm does not require a dynamic model of the robot, but just its mechanical constraints, and can be implemented in any industrial robot. The outcomes of the tests have been compared with both simulation and experimental results yielded by two trajectory planning algorithms taken from the literature.