X rays have been used for medical imaging since RÖNTGEN's fascinating discovery 125 years ago. The first radiographs of human hands were made public less than a month after his famous paper. The conventional X-ray sources integrated into the CT-machines of today's hospitals still rely on the same physical principles. X-ray imaging has traditionally offered high spatial resolution and low contrast for soft tissues such as the brain. Magnetic resonance imaging is therefore the method of choice for brain imaging in a clinical setting, although for cellular resolution studies it suffers from limited spatial resolution. The gold standard in post mortem brain imaging is histology, which involves fixation, embedding, physical sectioning, staining, and optical microscopy. Currently, section thickness limits isotropic voxel sizes to 20 µm. Advanced X-ray sources including synchrotron radiation facilities offer complementary modalities such as phase-contrast imaging and spatially resolved small-angle X-ray scattering. We showed that X-ray phase contrast of the human cerebellum with micrometer resolution yields complementary three-dimensional images to magnetic resonance microscopy with even better contrast and spatial resolution. Grating interferometry enabled us to visualize individual Purkinje cells in the nonstained cerebellum. Taking advantage of well-established paraffin embedding, Purkinje cells were visualized within the human cerebellum even with conventional instrumentation. Hard X-ray nano-holotomography allowed for label-free, three-dimensional neuroimaging beyond the optical limit with a spatial resolution below 100 nm. Spatially resolved smallangle X-ray scattering permitted the localization of periodic nanostructures such as myelin sheaths on square-inch brain slices and included the orientational information on the axons. These developments have contributed to the establishment of virtual histology and extended the conventional histology to the third dimension. Further advances are required to image the entire human brain with an isotropic micrometer resolution and to suitably handle the petabyte datasets.