An approach of calibration method for 3D sensing scheme based on shifted point array encoding is presented. Initially a 3D calibration benchmark is established in object space, the camera subsystem can be calibrated with the benchmark. The projection optics subsystem is bounded with camera subsystem in this sensing scheme so a constraint condition is imposed on both subsystems. Therefore it is possible to calibrated projection subsystem by introducing a concept of benchmark transmission. Once both subsystems have been calibrated, the 3D coordinates of range image can be obtained with moderate accuracy. The coordinate difference between measured data and prescribed benchmark is used to define a square-error based objective function, leading to a non-linear equation. By solving this equation, one is able to determine the required parameters of the system configuration. Experiment results show that, after proposed calibration procedure, the 3-D sensing scheme based on point array encoding is able to achieve such an accuracy as the standard deviation in X direction is 0.29mm, 0.24mm in Y direction, and 0.29mm in Z direction, respectively, for a measuring volume of 300×300×80mm3.