SAFT (Synthetic Aperture Focusing Technique) has substantial benefits over conventional UT and becomes increasingly popular due to the computational power available nowadays. In addition to more precise defect localization, also better separation of group indications and the improvement in grain induced (as well as electronically induced) signal-to-noise ratio (SNR) are important factors for utilizing SAFT. For practical application, it is not only important to ensure sufficient detection sensitivity but also to develop an economic solution. Besides system costs the inspection duration (respectively the throughput) is the key parameter to be optimized. Hence, the question for the optimum inspection grid size arises. Usually the scan grid for SAFT inspection is chosen empirically as this question has not been investigated sufficiently yet. The criterion that the sound beams shall cover the volume to inspect at least n-times – which is used for conventional ultrasonic inspection – is not helpful for SAFT. Of course, at least a single coverage has to be demanded, but tests have shown that this is not sufficient usually. Within this contribution the influence of the inspection grid is investigated in detail. The connection between the choice of the inspection grid and the appearance of artifacts is shown. By specifying a limit for the maximum tolerable artifact amplitude a criterion for the choice of the scan grid can be derived. The influence of specific inspection parameter, such as frequency, sound velocity, beam angle, and beam divergence is investigated systematically and it is derived, how the inspection grid can be optimized for SAFT inspection.