Machine vision systems need to cope with harsh environmental conditions, when used for guiding robots. To increase the reliability of these systems, they are either shielded from their environment or highly advanced, using sophisticated mathematical algorithms. In this paper the effects of external illumination on machine vision systems are investigated. For this the pose and the brightness of the external illumination and the internal illumination of the camera and its exposure time are varied. In a next step the correlation between the parameters of the external illumination and reliability of the machine vision system is analysed. Finally requirements are derived for further development of machine vision systems.