The determination of position and orientation errors made by large-scale cartesian coordinate robots using latest 3D-measurement methods has proved to be difficult in practise. A method has been developed to determine the mechanical inaccuracies of linear motions, perpendiculars, deflections, rotations and bends of a robot, relative to the common reference coordinates of the robot and the work area. The measurement method is based on the use of a laser diode and a position sensitive detector (PSD). A PSD is an optoelectronic sensor capable of providing position data about a light spot incident on its surface. The dual-axis, non-discrete PSD provides continuous analog X- and Y-axis information as the light spot transverses its active area. It senses the centroid of the light spot so that the position daa is indpendent of the focus of the spot. The PSD typically has an active area of - 1 cm - 20 cm , a resolution of 1/5000, nonlinearity of + 1% - + 15% and a fast response enabling accurate detection even of a rapidly moving light spot. The system developed consists of a laser diode transmitter, two beam splitters, a retro-reflective mirror, a PSD based receiver and a microcomputer unit for controlling the system and analysing the measured information. The divergence of the laser diode transmitter is - 0.1 mrad, it is small in size and lightweight (about 100 g), the direction of the laser beam of a fastened transmitter can easily be adjusted and it can be modulated electronically. The direction of the laser beam is not so sensitive to temperature variations as is for example an He-Ne-laser. The effective size of the PSD is enlarged with the aid of optics to have a diameter of 10 cm. Using a modulated laser beam the background light can be compensated. The position data from the PSD receiver to the memory of the microcomputer is gathered continuously and information is produced by a graphic printer in the form of graphs or numbers. This enables the measurement of the straightness of moving paths of machines such as robots, displacement and vibration measurements, optical position and angle sensing and surface flatness measurements. The performance of the system will be reviewed by considering the results of the practical measurements for determining mechanical misalignments of linear motions of cartesian coordinate robots.