The PAC-technique always claims to test the micro-surrounding of the probe atoms. Typically, the samples are macroscopic and more or less homogeneous and there is no debate about the usefulness of the method: substitutional sites, trapped vacancies or phase transitions are easily seen. Even the PAC-“fingerprint” of an amorphous material is known. In case of inhomogeneous samples, perhaps made out of different constituents, the question arises whether the PAC can contribute to the understanding of such materials or not. The article will show the different ways to introduce the probe atoms into the samples and discuss then the influence of these histories on the final site of the probe in the sample. In general, diffusion tends to place the probes into grain boundaries, whereas implantation reaches the bulk. This becomes important for nano-materials with their high fraction of internal surfaces. As a second, important difference for possible experiments the spatial distribution of the probes has to be considered. Implantation leads to a Gaussian shaped depth distribution of the probes. This corresponds – in a certain region – to a 3-dimensional distribution of probes in the sample, used e.g. when doping a semiconductor. In the production of special sensors (which apply e.g. the giant magnetoresistance (GMR) effect) one needs a different package, thin films (1-2 mono-layers). To apply PAC here, the probes have to be introduced during the fabrication of such a sensor. In these cases the probes have to be placed within a plane of nano-scale thickness.