Magnetic resonance imaging (MRI) has undergone a rapid development which is still continuing. In this article a survey is given of the present status of this new diagnostic tool in the evaluation of diseases of the central nervous system. When atoms with uneven numbers of protons or neutrons in a homogeneous magnetic field are tilted against the main vector of this field by a radiofrequency pulse, nuclear magnetic resonance can be observed. During the relaxation of the little dipoles back to the direction of the underlying magnetic field, a resonance signal is generated. The superposition of variable field gradients enables the scanning of sectional images in the axial, frontal and sagittal plane. The variables of H+-magnetic resonance which can be utilized for imaging are: the proton density, the relaxation times T1 (spin-lattice) and T2 (spin-spin) and flow effects. While the proton density in organic tissue fluctuates only by some 10%, the relaxation times may vary by several hundred per cent. Tissue contrast, therefore, is mainly based on relaxation times differences. The image character can also be influenced by variations of imaging parameters (i.e. repetition rate, interpulse delay, read out or echo delay) in different imaging sequences, such as the spin-echo and the inversion recovery technique. Depending on these imaging parameters T1 and T2 will contribute to the signal to a varying degree. This fact is most important for the diagnostic information of MRI. In initial clinical experiences in the diagnosis of diseases of the central nervous system, MRI has demonstrated high sensitivity in the detection of lesions (such as oedema, neoplasms, demyelinating disease), but less significance in lesion discrimination. In spinal disease the direct sagittal imaging of MRI enables MRI-myelography without contrast medium, superior to conventional myelography in many cases. For detailed evaluation of disc disease, however, the spatial resolution still has to be improved. Promising results have been obtained from flow effects. Depending on the flow velocity of blood, vessels appear white with intensive signals (slow flow) or black due to low signal intensities (rapid flow). MRI-angiography including measurement of blood flow seems possible. MRI-contrast media are not yet available for routine clinical use. Promising results have been reported on the basis of rare-earth elements, such as gadolinium Gd3+. These substances decrease T1 and T2 with subsequent increase in signal intensity. Concerning harmful side-effects of MRI, three possible sources have to be considered: the static magnetic field, the changing magnetic field, and radiofrequency heating.(ABSTRACT TRUNCATED AT 400 WORDS)