The incremental hole-drilling method is frequently used for residual stress depth distribution analyses, due to its fast and economical experimental execution. Depending on the planned use of the component, the drilled hole that is made to measure the residual stress can often be repaired or ignored if it does not affect the intended use of the part. Nevertheless an important experimental issue and assumption is the introduction of an ideal cylindrical hole into the component without additional plastic deformation. Although high-speed drilling is well established the consequences of the resulting hole geometries compared to ideal assumptions are not well known. Therefore, a detailed comparison between different bits and drilling techniques was carried out and is discussed in this paper in order to detect the best experimental conditions and to find out reasons especially for the lack of accuracy of the hole-drilling method for the first increments close to the specimens surface. It comes out that the orbital drilling with common used six-blade bits results in the best compromise of an ideal cylindrical hole and centricity to the center of the strain gage rosette. In the case of conventional drilling the hole geometry differs from the ideal one if six-blade bits were used due to the influence of chamfers at the cutting edges and a non 180° plane end face and also in the case of a two-blade bit due to a non 180° plane end face and the tendency to more eccentric holes. Diamond bits cannot be recommended under all tested conditions due to their geometrical undefined shape.