Microbeam radiation therapy (MRT) uses highly collimated, quasi-parallel arrays of X-ray microbeams of 50-600keV, produced by third generation synchrotron sources, such as the European Synchrotron Radiation Facility (ESRF), in France. The main advantages of highly brilliant synchrotron sources are an extremely high dose rate and very small beam divergence. High dose rates are necessary to deliver therapeutic doses in microscopic volumes, to avoid spreading of the microbeams by cardiosynchronous movement of the tissues. The minimal beam divergence results in the advantage of steeper dose gradients delivered to a tumor target, thus achieving a higher dose deposition in the target volume in fractions of seconds, with a sharper penumbra than that produced in conventional radiotherapy. MRT research over the past 20 years has yielded many results from preclinical trials based on different animal models, including mice, rats, piglets and rabbits. Typically, MRT uses arrays of narrow ( approximately 25-100 microm wide) microplanar beams separated by wider (100-400 microm centre-to-centre) microplanar spaces. The height of these microbeams typically varies from 1 to 100 mm, depending on the target and the desired preselected field size to be irradiated. Peak entrance doses of several hundreds of Gy are surprisingly well tolerated by normal tissues, up to approximately 2 yr after irradiation, and at the same time show a preferential damage of malignant tumor tissues; these effects of MRT have now been extensively studied over nearly two decades. More recently, some biological in vivo effects of synchrotron X-ray beams in the millimeter range (0.68-0.95 mm, centre-to-centre distances 1.2-4 mm), which may differ to some extent from those of microscopic beams, have been followed up to approximately 7 months after irradiation. Comparisons between broad-beam irradiation and MRT indicate a higher tumor control for the same sparing of normal tissue in the latter, even if a substantial fraction of tumor cells are not receiving a radiotoxic level of radiation. The hypothesis of a selective radiovulnerability of the tumor vasculature versus normal blood vessels by MRT, and of the cellular and molecular mechanisms involved remains under investigation. The paper highlights the history of MRT including salient biological findings after microbeam irradiation with emphasis on the vascular components and the tolerance of the central nervous system. Details on experimental and theoretical dosimetry of microbeams, core issues and possible therapeutic applications of MRT are presented.