With the rapid development in cell biology, particularly in the area of stem cells, the interest to explore the therapeutic potentials of stem cells has increased. In order to assess the therapeutic possibilities, the dynamics of cell migration and differentiation after implantation into the host organ must be studied. This again requires access to noninvasive imaging modalities allowing monitoring in individual subjects over time. During the last few years, various approaches using different imaging techniques, particularly optical imaging but also magnetic resonance imaging have been applied for this purpose. In order to detect the cells of interest against the host tissue background, the cells need to be labeled to produce a strong contrast. This has been achieved mostly with iron oxide nanoparticles, so-called USPIOs, which are incorporated into the cells, thus producing a strong signal loss in T2*-weighted MRI by virtue of susceptibility differences to the adjacent environment. There are basically three fundamentally different routes of cell labeling. First, the USPIOs are injected systemically and are consequently picked up by blood borne cells, in particularly macrophages. Using such a labeling approach, inflammation foci can be demarcated after the macrophages have invaded the inflammation areas. Secondly, cells can be labeled in vitro by using various techniques (lipofection, endocytosis, electroporation, etc.) to effectively incorporate the USPIOs into the cells. Such prelabeled cells are then implanted and their dynamics (migration, proliferation) are followed with T2*-weighted MRI. Finally, a thrid approach aims to stereotactically inject free iron-oxide label into the tissue area of interest where the label is expected to be incorporated by the cells in the close neighboorhood. When specific cells migrate out of this unspecifically labeled region (e.g. progenitor cells migrating from the subventricular zone to the olfactory bulb along the rostral migratory stream), this can be observed as a newly generated T2*-weighted contrast. The general idea of labeling cells with USPIOs requires, however, careful analysis and validation by independent techniques (mainly histology and immunohistochemistry) to minimize unambiguities and misinterpretations of seemingly cell-caused T2*-weighted contrast. There is a wide range of such confounding factors to be considered for image contrast assignment: vascular, BOLD based, macrophage activity, bleedings, pathophysiologically caused signal loss in the host tissue, transfer of contrast agents to host cells etc. To go beyond mere localization of labeled cells by MRI to allow for observation of functional fate, new strategies must be developed. For this purpose, two approaches have been followed recently. The first deals with responsive MRI contrast agents while the second approach involves the use of transgenic cell lines generating their own contrast mechanisms under cell-specific promoters. Both these strategies will, in principle, permit to detect when cells change from one functional cell state to another (pre-selected) new cell state because only then the responsive contrast agent will be activated or the intrinsic contrast mechanism up-regulated by activation of the corresponding promoter. Issues dealing with the potential and limits of all these different labeling strategies, their careful image interpretations and with issues involved of generating the contrast will be discussed. For this discussion, particular focus will be laid on application in experimental neuroscience and studies of cerebral diseases.